National Aeronautics and Space Administration

 

Landsat 7 Science Data Users Handbook
 

Foreword
The purpose of the Landsat program is to provide the world's scientists and application engineers with a continuing stream of remote sensing data for monitoring and managing the Earth's resources. Landsat 7 is the latest NASA satellite in a series that has produced an uninterrupted multispectral record of the Earth's land surface since 1972. Along with data acquisition and the USGS archival and distribution systems, the program includes the data processing techniques required to render the Landsat 7 data into a scientifically useful form. Special emphasis has been placed on periodically refreshing the global data archive, maintaining an accurate instrument calibration, providing data at reasonable prices, and creating a public domain level one processing system that creates high level products of superior quality. *Landsat Data is available for FREE -Download data via the USGS at: Glovis: http://glovis.usgs.gov/ Earth Explorer: http://edcsns17.cr.usgs.gov/EarthExplorer/ -Download data via GLCF at: GLCF: http://www.landcover.org/index.shtml The Landsat 7 Science Data User's Handbook is a living document prepared by the Landsat Project Science Office at NASA's Goddard Space Flight Center in Greenbelt, Maryland. Its purpose is to provide a basic understanding of the joint NASA/USGS Landsat 7 program and to serve as a comprehensive resource for the Landsat 7 spacecraft, its payload, the ground processing system, and methodologies for rendering Landsat 7 data into a form suitable for science.  

 

National Aeronautics and Space Administration

 

Landsat 7 Science Data Users Handbook
1.1 Background to the Landsat Program
The Landsat Program has provided over 38 years of calibrated high spatial resolution data of the Earth's surface to a broad and varied user community, including agribusiness, global change researchers, academia, state and local governments, commercial users, military, and the international community. Landsat images provide information meeting the broad and diverse needs of business, science, education, government, and national security. The mission of the Landsat Program is to provide repetitive acquisition of high-resolution multispectral data of the Earth's surface on a global basis. Landsat represents the only source of global, calibrated, high spatial resolution measurements of the Earth's surface that can be compared to previous data records. The data from the Landsat spacecraft constitute the longest record of the Earth's continental surfaces as seen from space. It is a record unmatched in quality, detail, coverage, and value. The Landsat platforms carry multiple remote sensor systems and data relay systems along with attitude-control and orbit-adjust subsystems, power supply, receivers for ground station commands and transmitters to send the data to ground receiving stations. The most recent Landsat mission, Landsat 7, offers these features: Data Continuity: Landsat 7 is the latest in a continuous series of land remote sensing satellites spanning 38 years. Global Survey Mission: Landsat 7 data will be acquired systematically to build and periodically refresh a global archive of sun-lit, substantially cloud-free images of the Earth's landmass. Affordable Data Products: Landsat 7 data products will be available through the EROS Data Center at the cost of fulfilling user requests (COFUR). Enhanced Calibration: Data from the ETM+ will be calibrated to better than 5% absolute, providing an on-orbit standard for other missions. Responsive Delivery: Automated request processing systems will provide products electronically within 48 hours of order. The continuation of the Landsat Program is an integral component of the U.S. Global Change Research Program. Landsat 7 is part of a global research program known as NASA's

 

Earth Sciences Enterprise, a long-term program that is studying changes in Earth's global environment. The goal of Earth Sciences Enterprise is to provide a better understanding of natural and man-made environmental changes. In the Landsat Program tradition, Landsat 7 will continue to provide critical information to those who characterize, monitor, manage, explore, and observe the land surfaces of the Earth over time. 1.1.1 Previous Missions Landsat satellites have been providing multispectral images of the Earth continuously since the early 1970's. A unique 38-year data record of the Earth's land surface now exists. This unique retrospective portrait of the Earth's surface has been used across disciplines to achieve improved understanding of the Earth's land surfaces and the impact of humans on the environment. Landsat data have been utilized in a variety of government, public, private, and national security applications. Examples include land and water management, global change research, oil and mineral exploration, agricultural yield forecasting, pollution monitoring, land surface change detection, and cartographic mapping. Landsat 7 is the latest satellite in this series. The first was launched in 1972 with two Earthviewing imagers - a return beam vidicon and an 80-meter multispectral scanner (MSS). Landsat 2 and 3, launched in 1975 and 1978 respectively, were configured similarly. In 1984, Landsat 4 was launched with the MSS and a new instrument called the Thematic Mapper (TM). Instrument upgrades included improved ground resolution (30 meters) and 3 new channels or bands. In addition to using an updated instrument, Landsat 4 made use of the multimission modular spacecraft (MMS), which replaced the Nimbus, based spacecraft design employed for Landsats 1-3. Landsat 5, a duplicate of 4, was launched in 1984 and even today after 26 years - 21 years beyond its 5-year design life - is still returning useful data. Landsat 6, equipped with a 15-meter panchromatic band, was lost immediately after launch in 1993. Table 1.1 lists key mission characteristics of the Landsat Program while Table 1.2 compares the sensors carried aboard these satellites. A detailed Landsat Program Chronology is also available.

 

Table 1.2 Landsat Satellites and Sensors   4  .1 Landsat Mission Characteristics   Table 1.

With the passage of Public Law 98-365. and operations. 1985. NOAA was to pursue procurement of future remote sensing Landsat products and services from the private sector. a joint venture between RCA and Hughes Aircraft. NOAA was directed to delegate management of the Landsat 4 and 5 satellites and their data distribution to the private sector. national security. In addition. now called Space Imaging Corporation. In 1985. international obligations and public safety. NOAA solicited bids to manage the existing Landsat satellites and to build and operate future systems.1. development. The implementation of this policy required the transfer of government-developed space technology to the private sector in such a manner as to protect its commercial value. EOSAT retained exclusive sales rights to all Landsat 4 and 5 Thematic   5  . won the competitive bidding process in August 1984 and took over operation of the Landsat system on September 27. the "Land Remote Sensing Commercialization Act of 1984". Government agencies. including NASA and NOAA.S. U. The Earth Observation Satellite Company (EOSAT).2 The EOSAT Era In the mid 1980's. were directed to attain their commercial space objectives without the use of direct federal funding by entering into appropriate cooperative agreements with private sector corporate entities to encourage and advance private sector basic research.1. Commercial sector space activities developed under this mandate were to be supervised or regulated by federal agencies only to the extent required by law. From 1985 to 1994. which included retention of technical data rights by the private sector.

Landsat 6. In response. into law in September of that year.  Returned management of the Program to the Government. This agreement between Landsat Program management and EOSAT Corporation on cost and reproduction rights for Landsat 4 and 5 Thematic Mapper data remains in effect and was last updated in October 1996. program management strategies and implementation guidelines.1.   6  . the U.3 Basis in Law for Landsat By 1992.R. it had become clear that the high cost of commercially provided Landsat data had greatly restricted its use in research and other public sector applications. recent legislation relevant to Landsat: . EOSAT also won competition to produce the next satellite in the series.  Foster development of advanced land remote sensing systems and opportunities for commercialization.  Established a data policy of distributing Landsat data at the cost of fulfilling a user request (COFUR). Subsequent NASA and NOAA memoranda later that summer brought the current Landsat 7 mission into existence. 1. 1994 Presidential Decision Directive (NSTC-3) defined the new Landsat 7 data policy.HR1278 Short title: "National Oceanic and Atmospheric Administration Authorization Act of 1997” .HR1702 Short title: "Commercial Space Act of 1997" Visit Recent and Pending Legislation Affecting the Landsat Program for detailed public law information. 6133. Congress passed H. The loss of Landsat 6 in October 1993 suddenly made the new Landsat 7 mission imperative. the "Land Remote Sensing Policy Act of 1992".  Directed that preliminary work begin on a new Landsat 7.S. This law established a new national land remote sensing policy which:  Abandoned full commercialization of the Landsat Program.Mapper (TM) data until July 1994. Other. Fiscal Years 1998 and 1999" .HR1275 Short title: "Civilian Space Authorization Act. at which time Landsat data over ten years old became available from the National Archive at the EROS Data Center (EDC). A May 5.

high quality visible and infrared images of all landmass and near-coastal areas on the Earth. NASA was also responsible for the satellite launch and performing a 60-day in-orbit check out before handing operations to the USGS.1. coverage and spectral characteristics to allow comparison for global and regional change detection and characterization. The following web sites should be visited for additional information: NASA Landsat 7 Project developed the Landsat 7 System. foreign. and testing the Landsat 7 spacecraft. ETM+ instrument. archiving.   7  . Data input into the system will be sufficiently consistent with currently archived data in terms of acquisition geometry. The USGS is responsible for operation and maintenance of the satellite and the ground system for the life of the satellite. continually refreshing an existing Landsat database.S. from NASA/USAF/USGS to NASA/NOAA/USGS to a bi-agency NASA/USGS partnership. implements mission policies. civil. In this role the USGS captures. In this capacity EDC directs on-orbit flight operations. developing. Specifically.1. it meant designing. and private sector use as well as academic. and interacts with International Ground Stations.4 System Operation and Management The Landsat 7 Program management structure changed repeatedly from 1992 through 1998. and the end-to-end ground system. processes. Another goal of the project is to expand the uses of such data. EDC also provides a public interface into the archive for data search and ordering and handles billing and accounting procedures. national security. directs acquisition strategy. calibration. The Landsat 7 project will continue to make Landsat data available for U. The USGS Center for Earth Resources Observation and Science (EROS) manages the overall Landsat 7 Mission Operations. NASA is still responsible for verifying data processing integrity and assuring high image quality. and distributes the data and is responsible for maintaining the Landsat 7 data archive.2 Landsat 7 Mission Objectives The Landsat 7 Mission Objective is to provide timely. EDC captures Landsat 7 data and performs pre-processing. and commercial uses. and distribution functions. NASA is responsible for the development and launch of the Landsat 7 satellite and the development of the ground system. The Landsat Project at Goddard Space Flight Center manages these responsibilities with Hughes Santa Barbara Remote Sensing building the sensor and Lockheed Martin Missiles and Space developing the spacecraft. As described in the Landsat 7 Management Plan. product generation. 1.

substantially cloud free.1 Overall Mission Objectives Landsat 7 is to have a design lifetime of five years.  Accept data acquisition and product requests from users. orbit in tandem with EOS-AM1 for near coincident observations. The overall objectives of the Landsat 7 Mission are:  Provide data continuity with Landsats 4 and 5. international and commercial communities.  Supply data to users at the cost of fulfilling the user request. capture.)  1. land images.2.  Produce the equivalent of 100 Level 0R products and 100 Level 1 products per day.  Produce browse and metadata for all full and partial scenes acquired. (In particular.2.           8  .  Support Government.1.  Provide for rapid turnaround of priority acquisitions and processing.  Play a vital role in NASA's Earth Observing System (EOS) by promoting interdisciplinary research via synergism with other EOS observations. and archive the equivalent of 250 ETM+ scenes per day.2 Specific Performance Requirements Some specific requirements for the Landsat 7 System include the following:  Acquire. with phased expansion capabilities to handle an increased processing load.  Provide communications downlinks for data capture by fixed and transportable X-band ground stations using the Consultative Committee for Space Data Systems (CCSDS) standard protocol for communication of data.  Offer 16-day repetitive Earth coverage.  Build and periodically refresh a global archive of Sun-lit.  Make data widely available for the cost of fulfilling a user request (COFUR).

3 System Capabilities The Landsat 7 system design can best be described as robust. However.1. substantially cloud-free images of the Earth’s entire land surface.S. Landsat 7 will be able to image the Earth's landmass systematically every 16 days.  Provides improved access to International Ground Station data. New and unusual system capabilities include:  Provides for a systematic collection of global.  Provides for a high volume of data collection (averaging 250 scenes per day into the U. A "Long Term Plan" has being developed to define the acquisition pattern for the Landsat 7 mission in order to create and periodically update this global archive.1 Global Survey Mission An important operational strategy of the Landsat 7 mission is to establish and maintain a global survey data archive.  Provides data available for ordering by end users within 24 hours of capture at EDC. high resolution.   9  .3. archive). following the same "Worldwide Reference System" used for Landsats 4 and 5. unlike previous Landsat missions. multispectral data. 1.  Provides delivery of up to 100 scene products per day.  Uses cloud cover predicts to avoid acquiring unusable data. Landsat 7 will endeavor to systematically capture sun-lit. See Chapter 5 for descriptions of the Worldwide Reference System and Long Term Acquisition Plan.

Product media options include Exabyte tape (8 mm 8200. Users of Landsat 7 data will have improved access to this data.3. 1.3.1 . New users should contact the EDC-DAAC User Services Office (605-594-6116 voice. 8mm 8500). Landsat 7 product orders may be placed electronically via the web using the USGS Global Visualization Viewer or the Global Land Cover Facility. CD-ROM or electronic transfer via FTP.3 Enhanced IGS Access Imagery of foreign landmasses will be recorded and down linked to EDC yet the temporal depth will be a fraction of what's available at the international ground stations. Email for assistance in setting up data purchase accounts and arranging payment method in advance. 605-594-6963 fax. each international ground station collecting Landsat 7 data is required to send periodic inventory information in the form of scene metadata to the LP  10  .2 Rapid Data Availability Figure 1.Landsat 7 Data Distribution System The Landsat 7 data distribution system will provide access to Landsat 7 Level 0R data products within 24 hours of collection and Level 1 processed products within 48 hours of request. Billing and accounting are handled via ECS-registered prepaid accounts. As stipulated in the NOAA-IGS Memorandum of Understanding.1.

tracking and command for narrowband communications. S-Band is used for commanding and housekeeping telemetry operations while X-Band is used for instrument data downlink. and torque rods and magnetometers for momentum unloading.8 MB Quicktime movie. The IGSs also have the option to send browse imagery to the LP-DAAC for viewing by interested users. a static Earth sensor. standard and as such. A 378 gigabit Solid State Recorder (SSR) can hold 42 minutes of instrument data and 29 hours of housekeeping telemetry concurrently. The injected spacecraft. archive. and 12 one pound-thrust jets.1 .184 square feet) and two 50 amp-hour Figure 2. roll. Users will discover on EDC's web servers internet links to IGS browse systems and information on IGS product types and ordering protocols. sun synchronous. yaw. and hydrazine   11  . Power is provided by a single Sun-tracking solar array (four 74" by 89. 1750A processor. on-board computing for power management.DAAC.S. skewed). attitude control.S.Landsat 7 Launch Nickel-Hydrogen batteries. Spacecraft weight is approximately 4632 pounds at launch. attitude control. They can only be discovered and then ordered directly from the IGS data distributor(s). NASA standard telemetry. and store and forward wideband data. No IGS products will be archived at EDC. three axis stabilized.3" panels . 1999. 2. IGS metadata is structured according to the U. The LP-DAAC will also support access to foreign products in other ways. the Enhanced Thematic Mapper Plus (ETM+).1 Overview The Landsat 7 satellite was successfully launched from Vandenburg Air Force Base on April 15. is available for web-based searching using tools developed for the U. direct energy power transfer.2 Platform Details The satellite provides active. is a 5000 poundclass satellite designed for a 705 km.2 and in this 1. satellite commanding and failsafe protection management. three 2-channel gyros with celestial drift updating. 2. Orbit control and backup momentum unloading is provided through a blow-down monopropellant hydrazine system with a single tank containing 270 pounds of hydrazine. earth mapping orbit with a 16-day repeat cycle. associated plumbing. Attitude control is provided through four reaction wheels (pitch. including a single solar array with multiple panels. Landsat unique direct. with on-board power storage. depicted in Figure 2. Its payload is a single nadirpointing instrument. The Delta II launch vehicle left the pad at 11:32 PDT and performed flawlessly.

which is the collection of housekeeping and satellite processor reports. These include attitude control. The Satellite Segment also includes the Aerospace Ground Equipment (AGE) which is composed of all the electrical (EAGE). The satellite attitude control uses precision mode. is used to verify the readiness of the satellite during pre-launch testing and during launch operations and readiness reviews. and an orbit adjust maneuver mode. which is a combination of stellar and inertial guidance sensors to maintain the spacecraft platform within 0. Also. an on-board solid-state recorder is used to store imagery of foreign landmasses for subsequent transmission to EDC. it consists of support system test functions and protection of the satellite during transport. solar array pointing maintenance. electrical power management. The on-board processor performs autonomously executed functions for wideband communications. and stored command execution. Additionally.The Landsat 7 satellite as viewed from the and stored commands. Narrowband communications include processing of real-time Figure 2. local vertical acquisition mode.   12  . switchable to three steerable antennas that have a downlink beam width of 1. yaw gyro backup mode.reaction control for momentum compensation and orbit adjust maneuvers. One particular type of EAGE. processing and authentication of command messages and transmission of telemetry data. redundancy management. antenna steering. sun side. thermal profile maintenance. and satellite control. mechanical (MAGE). Wideband communications for payload data transmission to the ground incorporates four X-band transmitters. the launch site equipment (LSE).2 . instrument and propellant ground equipment and related software used to integrate the satellite to the launch vehicle. The capability exists for an initial attitude control rate-nulling mode.2 degrees.015 degrees of earth pointing. battery management.

Short-Term Planning The objectives of short-term planning are to schedule communication contacts for telemetry. and project elements. Daily Scheduling On a daily basis. The MOC ingests the long-term plan prior to launch and uses the data as the main database from which to plan daily schedules of instrument and SSR activities. and to generate daily reports summarizing the disposition of imaging requests and time-ordered scheduled ETM+ imaging events for the latest 48-hour period.2. To assist in the scheduling process. After the FOT has iterated through the scheduling function and resolved all conflicts. the MOC will pass the 48-hour schedule to the load generation function for final compilation. to include special requests into the scheduling process. planned satellite maneuvers. and metadata on acquired or archived imagery from the EDC-DAAC. and X-band downlink services from which the MOC will generate an absolute time stored command load. short-term planning. tracking and commanding services on the Landsat Ground Network (LGN). and daily scheduling. Long-Term Planning The prime goal of the Landsat 7 mission is to refresh the global archive.3.   13  . to provide ample time for coordination with the science community. the MOC must generate a nominal 48-hour set of imaging. SSR or other resource availability.1). The long-term acquisition plan was generated early. the MOC receives planning aids from the FDF (Table 2.1 Scheduling Planning and scheduling takes place in three categories. long-term planning. including global refresh requirements. request priority. international resources. The MOC will schedule the 48-hour set of activities based on a number of criteria. and international imaging needs. SSR activities. and cloud cover predictions.3 Functions 2. Because the Landsat 7 mission orbit profile operates on a repetitive 16-day cycle. The longterm plan takes into account the calibration plan. a cloud cover prediction from the National Meteorological Center (NMC). program management. the global archive refresh strategy was designed years before the Landsat 7 satellite launch.

Telemetry data is generated and recorded at all times and contains all of the information required to monitor and assess the health of the satellite. formats.Table 2. The zenith antenna is used for Tracking Data and Relay Satellite (TDRS) communications. The S-band communications are conducted through two omni S-band antennas located on opposite sides of the spacecraft (nadir and zenith pointing). the nadir antenna is used for LGN communications. verify day-to-day operations.1 MOC/FDF Product Exchange Product from MOC Telemetry and displays Orbit maneuver data Post maneuver evaluation IGS position data Maneuver planning Product to MOC Post maneuver evaluation and report Orbit State Vectors Planning Aids Local Oscillator Frequency (LOF) Star Catalog Maneuver planning Maneuver plan Orbit maneuver parameters Station Acquisition data Frequency Realtime during each pass. and assist in anomaly resolution.2 Tracking and Spacecraft Control The radio frequency (RF) communications system provides S-band (narrowband) telemetry for housekeeping data and tracking ability. The on-board telemetry data formatter continually collects. TBD after receipt of playback As needed (within 30 minutes of maneuver command) As needed within 48 hours of maneuver completion As needed (within 24 hours of data receipt) Daily Weekly Monthly As needed As needed 48 hours before maneuver 48 hours before maneuver Daily 2. and outputs a telemetry data stream into a CCSDS compliant downlink protocol   14  . Each antenna provides essentially hemispherical coverage.3.

and  Inclination Maintenance. the new orbit state vectors table is uplinked to a scratch buffer.3. The semi-major axis is biased high and permitted to decay over time. The operators perform the inclination maneuver to keep the satellite-descending   15  . The MOC receives tracking telemetry on a daily basis that shows the position and velocity of the spacecraft. The In-plane maintenance maneuver. During daily scheduling. The bias applied to the orbit varies with the amount of environmental drag. The encoded data may also be sent to the S-band transponder for transmission to the MOC via a TDRS or LGN ground station.90 degrees. maintains the semi-major axis within an acceptable tolerance of the mission orbit semi-major axis.3 Orbit Maneuvers During the Landsat 7 mission. The orbit state vectors must provide an attitude accuracy of 375 meters at 40 hours and must be uplinked daily in order to maintain the satellite within mission parameters. which is a function of solar activity. The MOC receives the orbit state vectors file from the FDF and validates the information. stored commands deactivate the active orbit state vectors table. The Flight Dynamics Facility (FDF) is responsible for computing the actual spacecraft position and velocity for the last 61 hours and generating predicted orbit state vectors for the next 72 hours. After the MOC operators uplink and activate the absolute stored command load. and reactivate the active orbit state vectors table. The inclination maintenance maneuver involves a yaw slew by approximately +/.and sends the encoded data to the SSR for recording. the MOC operators perform two types of orbit maneuvers to ensure the satellite's orbit remains within the Landsat 7 mission parameters. move the orbit state vectors from the scratch buffer into active memory. At the appropriate epoch time. the Flight Operations Team (FOT) inserts an orbit state vectors activation command into the daily absolute command sequence. The new orbit state vectors are compared against the old orbit state vectors in the same way the flight software makes this comparison on the satellite after the receipt of the new orbit state vectors load. The two types of orbit maneuvers are:  In-plane Maintenance. also called a drag make-up maneuver. The MOC converts the orbit state vectors data into a Landsat 7 orbit state vectors table. 2. Onboard software interpolates this new data to generate the positional information contained in the Payload Correction Data (PCD).

the MOC conducts TDRS contacts for acquiring tracking data. The MOC generates orbit state vector loads for each set and make them available to the FOT. The FOT mission planner will insert an orbit state vector activation command that is timed to execute at the end of the maneuver window. and generates new orbit state vectors for routine delivery on the day following the burn. verifies the success of the orbit adjust. Additionally. The MOC uses the MOC-resident propulsion model to calculate the remaining propellant on-board the satellite and analyze the propulsion system. Following the orbit adjust. The FDF evaluates the performance of the satellite components used during the orbit adjust and provide that information to the MOC to update the MOC-resident propulsion model. On the day of the orbit maneuver. the FOT schedules TDRS tracking passes to occur immediately after the maneuver. The FDF receives tracking data from the space network and Landsat telemetry data from the MOC that are necessary to monitor the orbit and informs the MOC when the orbit has been perturbed sufficiently to require an orbit adjust. The FOT coordinates the orbit maneuver times with the NCC for TDRSs support and with the NOAA SOCC for CDAS support to facilitate the communications support needed to monitor the orbit maneuver in real-time. The Flight Operations Manager (FOM) reviews the data and the resulting commanding prior to merging the commanding into the absolute time command load for the day of the orbit maneuver. Once the FOM makes the final decision to perform the orbit maneuver. The FDF is responsible for monitoring the satellite orbit to ensure that its orbit remains within mission parameters. and passes the command data and a maneuver plan to the MOC. the FOT uplinks the absolute time stored command load to the satellite.node within a 30-minute box (09:45 to 10:15 AM). Sometime after the FOT activates the stored command load and before the orbit maneuver occurs. At that time. using this tracking data. including an orbit maneuver time window. The FDF calculates the orbit maneuver command data. The orbit state vectors load remains in the on-board scratch buffer until after the burn takes place.   16  . the FOT uplinks the appropriate orbit state vectors load. The FDF calculates the new satellite orbit. the FDF sends the MOC two sets of orbit state vectors: one to execute in a burn scenario and one to execute if the MOC cancels the orbit maneuver. The MOC ingests the orbit maneuver command data and generates a set of absolute time commands. the FOT executes real-time commands to activate the post-burn orbit state vectors by copying the orbit state vectors from the scratch buffer to the active orbit state vectors table and starting ephemeris processing of the new orbit state vectors.

operated by the Space Network.4 Communication Links Ground sites exist at Sioux Falls.2. Wallops Virginia (WPS). Alaska (Alaska Ground Station.S. South Dakota (The Landsat Ground Station. The SSR can hold up to 42 minutes (approximately 100 scenes) of data at 150   17  . 2. The figure below provides a view of the ground stations' acquisition circles. The SSR accepts two inputs at 75 Mbps. Ground Station Acquisition Circles In addition to the ground sites. All ground sites are equipped with 11-meter antennas. AGS. SGS and AGS are also known as the EOS Polar Ground Sites (EPGS) and are managed by Wallops Flight Facility (WFF). NASA code 530 is utilized.5 Solid State Recorder The SSR is used to capture wideband data from the ETM+. or AGS). and Svalbard. The NASCOM division of NASA provides all network lines.3 .U. TDRS is also used tracking and data collection for FDF generation of Landsat 7 spacecraft ephemeris. which is subsequently uploaded. and LGS are capable of receiving both S-band (housekeeping) at a downlink rate of 4 Kbps and X-band (payload) data simultaneously. Poker Flat. or LGS). Figure 2. the TDRS. TDRS enables downlink of real-time and stored S-band data and command of the spacecraft. Norway (SGS). for use in precision attitude control. SGS.

CADUs are recorded in the same order as received from the ETM+. the two 75 Mbps bitstreams are read out of memory and sent to the broadband switching unit. In this case. During playback.864Kbps and plays back stored telemetry data at 256 Kbps to the S-Band transponder. The SSR records and plays back wideband data in numbered logical blocks.   18  .Mbps.914 Mbps. each corresponding to an ETM+ collection interval. Partial CADUs may be recorded if the ETM+ collection interval extends beyond the commanded SSR record interval. Narrowband data is captured from by the SSR from the Telemetry Data Formatter. The bitstreams include the CADUs generated by the ETM+. S-Band telemetry data is stored separately from wideband image data and can be recorded during load shedding. which are used by the MOC in commanding the recorder. Reed-Solomon encoding is performed on record while Reed-Solomon decoding is performed on playback to recover data from possible dynamic RAM list errors. Each subinterval includes all of the CADU data required to process the subinterval as a separate ETM+ collection. if the ETM+ is turned off before the end of the SSR data recording area is achieved. As a result. The SSR accepts two input rates of 1. The SSR is capable of either recording or playing back wideband data (but not both simultaneously) while simultaneously recording and/or playing back narrowband telemetry data. Record intervals.216 Kbps or 4. may be subdivided for playback if more than one scene is collected. or as a result of a ground command to disable wideband recording. The SSR contains error detection and correction capability to preserve the integrity of the stored wideband and narrowband data. which consists of one or more Landsat scenes. individual subintervals may contain partial CADUs. A second pair of 75 Mbps bitstreams can also be played back for a total aggregate rate of 300 Mbps. The SSR records ETM+ channel access data units (CADU) as two bitstreams. each at a nominal rate of 74. each resulting subinterval is defined such that data in the vicinity of each subinterval boundary are included (redundantly) with both subintervals.

sunsynchronous and circular orbit at a 705 km nominal altitude.2 degrees. before the solid-state detectors at the focal plane collect it.1 Sensor Overview Landsat 7's sensor . with an orbit inclination of 98. and the addition of two solar calibrators. The sensor is a derivative of the Thematic Mapper (TM) engineered for Landsats 4 and 5. 3. The bidirectional scan mirror assembly (SMA) sweeps the detector's line of sight in west-to-east and east-to-west directions across track. but is more closely related to the Enhanced Thematic Mapper (ETM) lost during the Landsat 6 failure. a 3 axis stabilized spacecraft located in a near polar. the improved spatial resolution for the thermal band.2. The primary performance related changes of the ETM+ over the TM's are the addition of the panchromatic band Figure 3.2 ETM+ Design Scene energy passes through a number of major ETM+ subsystems. The ETM+ design provides for a nadir-viewing. A Ritchey-Chretien telescope focuses the energy onto a pair of motion compensation   19  .3. depicted in Figure 3.1 . The ETM+ is designed to collect. while the spacecraft's orbital path provides the north-south motion. eight-band multispectral scanning radiometer capable of providing high-resolution image information of the Earth's surface when operated from Landsat 7.the Enhanced Thematic Mapper Plus (ETM+) was built by SBRS. filter and detect radiation from the Earth in a swath 185 km wide as it passes overhead and provides the necessary cross-track scanning motion while the spacecraft orbital motion provides an along-track scan.Enhanced Thematic Mapper Plus and two gain ranges (added for Landsat 6).

The spectral filters for the bands are located directly in front of the detectors.   Figure 3.1 Detector Geometry Figure 3.ETM+ Optical Path The aligned energy encounters the Primary Focal Plane (PFP). This arrangement provides for a contiguous bank of 32. scan line corrector) where it is redirected to the focal planes.2.e. 16. and 8 detectors for band 8. A portion of the scene energy is redirected from the PFP by the relay optics to the Cold Focal Plane where the detectors for bands 5.mirrors (i. 3. The even-numbered detectors are arranged in a row normal to the scan direction while the odd-numbered detectors are arranged in a parallel row. where the silicon detectors for bands 1-4 and 8 (panchromatic) are located. The scan line corrector is required due to the compound effect of along-track orbital motion and crosstrack scanning which leads to significant overlap and underlap in ground coverage between successive scans. 7. The temperature of the cold focal plane is maintained at a predetermined temperature of 91° K using a radiative cooler.2 .3 illustrates the relative position of all the detectors from both focal planes with respect to their actual ground projection geometry. bands 1-5   20  . off exactly one IFOV in the along scan direction. and 6 are located.

5µ IFOVs which translates to 4 15 meter samples. Band Offsets During a scan the actual ground observed by each band's detectors is not identical due to the horizontal spacing of detector rows within and between bands. The band 8 odd and even detectors are sampled simultaneously.3 .5 microradian IFOVs. twice per minor frame (i.3 one should note the spacing between bands as measured in 30 meter 42.   Figure 3. The registration offsets for the odd and even detectors will therefore always differ by four samples for both forward and reverse scans.e.and 7. which covers a ground swath approximately 185 kilometers wide. and 8 data lines at a time) of along track image data is added to the acquired subinterval. these numbers represent a band's unique leading edge preamble that occurs before coincident data is collected by a band's forward or reverse focal plane neighbor. The detector arrays are swept left to right (forward) and right to left (reverse) by the scan mirror. these numbers represent the first order zero fill offsets that LPS uses during 0R processing to achieve image registration at the level 0 level. one sample). With each sweep or scan an additional 480 meters (32. Other factors such as detector offsets within a band and sample timing must be considered to calculate registration offsets accurately. Taken individually.5 42. This seemingly   21  . 16. and band 6 respectively. Again referring to the ground projection illustration in figure 3.5µ IFOVs. The detector rows within bands 1-5 and 7 are separated by 2. Taken cumulatively.Detector Projection at the Prime Focal Plane Detector Offsets Band 8 detectors rows are separated by 2 42.

2 Registration Offsets Over the years. digitizes. The six minor frames immediately following the SLS minor frame describe the spacecraft time code as illustrated in Figure 3. The time code data preempts all minor frame video except band 6 data. The registration offsets for forward and reverse scans will always differ by these amounts. A major frame contains the data for an entire period of one complete scan of the ETM+ scan mirror. in alternating minor frames. These 8 bit buffers are 3300. A 2 band 6 IFOV odd-to-even detector spacing is realized on forward scans while a 3 band 6 IFOV spacing occurs on reverse scans. different ground system engineers have characterized Landsat focal plane offsets in different ways that resulted in negative and positive offsets depending upon the forward and reverse scan directions and origin of the image grid. and 13.2.5 IFOV later than the odd detectors within a minor frame of data. and groups analog video signals from the ETM+ scanner to form scene data. however.200 elements in size for the 60. The band 6 data alternates between the odd and even detectors for each successive minor frame and are synchronized to odd pixel data for the   22  .5µ IFOVs which translates to 2. The delay effectively separates the odd and even detectors an integral multiple of IFOV's apart in sampling space. The band 6 odd and even detectors are separated by 5 42.2. The registration offsets for forward and reverse scans will always differ by these amounts. A major frame is partitioned into minor frames .5 band 6 samples. The time code pattern preempts all video data except band 6 although the band 6 data for the first minor frame is invalid. A 2 IFOV odd-to-even detector spacing is realized on forward scans while 3 IFOV spacing occurs on reverse scans. The line sync code is generated by the AEM at the beginning of each new scan line and inserted into minor frame zero which is also called scan line start (SLS). The minor frame data structure is 85 words (8 bits) in length consisting of 16 separate groups of 5 words. For Landsat 7 we have declared all shifts as positive from column 1 in the 0R image buffers. The odd and evens are sampled. 3. which separate the odd and even detectors. 30.curious design makes sense because the multiplexer samples the even detectors . 4 data words from band 6 and a one spare. 6600. This pattern is based on the architecture of the Landsat 7 auxiliary Electronics module (AEM) that samples. and 15-meter bands respectively.4.a specific pattern for organizing groups of ETM+ data words. an integral multiple of IFOVs.3 ETM+ Sample Timing The ETM+ data stream is composed of a continuous succession of major frames. 3.

The first valid band 6 data are from the even detectors and occur in minor frame two.     23  . For reference.5.ETM+ Sample Timing .4 and 3. Figure 3. 6313 minor frames of scene data are nominally transmitted during any given scan cycle.4 . which is mechanically/optically. Bands 1-5 are allocated to format 1 while bands 7 and 8 are allocated to format 2. The digitized scene data can be organized into either of two minor frame scene data formats as depicted in Figures 3. The two band 6 data streams allocated to formats 1 and 2 are low gain and high gain respectively.first minor frame. triggered.Format 1   Scene data transmission for the other bands starts at the minor frame boundary immediately following the time code and continues until start of the next end of line pattern code.

1 Scan Mirror Assembly Table 3.3.050 in x 16.719 milliseconds 4.743 milliseconds 10.21095 9.925 milliseconds 6.ETM+ Sample Timing .997 Hz 60.695 degrees 142.42191 rad/sec 2.611 microseconds 6320 IFOVs < 9.1 Swath width at 0 degrees North Swath width at 40 degrees North Scan length at 0 degrees North Scan length at 40 degrees North Active scan amplitude Scan period Scan Frequency Active scan time Turn around time Object plane scan rate Mirror scan rate 30 m IFOV dwell time Scan line length (excludes turn around) Inertia Clear aperture 185 km 187 km 480 m 484 m 7.Format 2     3.3 ETM+ Subsystems 3.4 in-lb-(sec*sec) 21.250 in   24  .Figure 3.5 .

  25  . a torquer. Additionally.743 µs. which has a very low coefficient of thermal expansion and thus eliminates problems due to thermal expansion. SME 1 and SME 2. However. Both the tube-like central baffle and the outer housing have a series of annular baffles for stray light control. both SMEs have a primary SAM mode of operation and a back up Bumper Mode of operation. There are redundant sets of scan mirror electronics.9 MHz system clock by circuits in whichever multiplexer is providing the system timing.3. The SMA consists of a flat mirror supported by flex pivots on each side (which have compensators to equalize pivot reaction torque). The telescope structure is constructed using a graphite-epoxy laminate. The resulting active scan in each direction is closely controlled to 60.8 cm) 6. the graphite-epoxy laminate is hygroscopic and can change dimensions due to moisture absorption.2 Telescope Table 3.The Scan Mirror Assembly (SMA) provides the crosstrack scanning motion to develop the 185 km long scene swath. SMA characteristics are listed in Table 3. The SMA derives timing from a 10 MHz clock routed by buffers in the multiplexer. The telescope's characteristics are summarized in Table 3. The motion of the mirror in each direction is stopped by the bumper.64 cm) 6. SAM mirror angle pulses are used by the multiplexer to synchronize the detected scene data.56 in (16.2 Primary mirror clear aperture outer diameter Primary mirror clear aperture inner diameter Telescope effective clear aperture Effective focal length f/# Mirror material Mirror coating 16. 3.0 Ultra-low expansion (ULE) glass Enhanced silver The telescope is a Ritchey-Chretien configuration with a primary and secondary mirror and baffles.2.0 in (40.66 cm) 1020 square cm 96 in (243. SME 2 is an identical back up electronic package to SME 1.1. The central baffle is made from aluminum. The 10 MHz is counted down from the 74. The amount of torque applied is controlled by the SME microprocessor as determined from the SAM mirror angle pulses. 2 leaf spring bumpers and scan mirror electronics (SME). a scan angle monitor (SAM). and is boosted by precision torque pulses during the turnaround period. Care was exercised during integration and test to minimize moisture absorption and to maintain knowledge of the telescope's condition.

  Figure 3. which has primary and redundant tachometers.6 .3 Scan Line Corrector The Scan Line Corrector (SLC) is an electro-optical mechanism composed of two parallel nickel-plated beryllium mirrors set at an angle on a shaft. The primary and redundant SLC drivers are implemented on the MEM A1 and A2 circuit cards respectively.6). rotates about an axis normal to the axis of the scan mirror in a saw tooth fashion.3. driven by a torquer. The SLC driver produces a linear saw tooth   26  .3. A rectilinear scan pattern is produced using the SLC instead of the zigzag pattern that would be produced without it (Figure 3.6 mr/sec. The SLC is positioned behind the primary optics and compensates for the along track motion of the spacecraft that occurs during an active SMA cross track scan.ETM+ Scan Line Correction A feedback control system is employed in the SLC drive electronics to achieve a constant angular rate of 576. The shaft.

The system focus is optimized for the Pan Band.00204 in x 0.056 in Linear image displacement rate 9.6 µrad/sec SLC linear scan angle 35. On top of each stack is a cylindrical black radiative cooling tower to help dissipate the heat from the preamplifiers. Table 3.7 µrad Mirror separation 1.99 Hz Scan period 71.00177 in in Detector Area IFOV size 42. The preamplifiers are mounted on the Prime Focal Plane Assembly.25 µrad x 18.3. The array for the Pan Band contains 32 detectors also divided into odd-even rows. The SLC driver receives a 208 kHz clock and a buffered line start signal from the Timing Control Subsystem. Table 3.5 µrad Center to center spacing along track 0.3 lists the SLC characteristics. Table 3. and consist of two stacks of flat hybrid modules.signal with a fast fly back.462 ms Scan rate in object space 9.22 in/sec Mirror material Nickel plated beryllium Mirror coating Enhance silver 3. The Prime Focal Plane Array is a monolithic silicon focal plane made up of five detector arrays: Band 1 through Band 4 and the Pan Band. The Prime Focal Plane Array is located at the focal plane (surface) of the telescope. which has the highest spatial resolution. the Prime Focal Plane Array and the two preamplifier stacks.4.600 in Linear image displacement amplitude .00408 in x 0. Its characteristics are listed in Table 3.02 µrad Linear scan angle in object space 583.4 Prime Focal Plane Assembly Design Parameters Parameter Bands 1 through 4 Pan Band Number of detectors 16 per band 32 0.00408 Detector Size 0.5 µrad 21.610 µrad/sec SLC rotation rate 576.00408 in 0. The arrays for Bands 1 through 4 contain 16 detectors divided into odd-even rows.4 Primary Focal Plane The Prime Focal Plane Assembly consists of three major subassemblies.00204 in Center to center spacing between 0.01020 in 0.3 Scan Line Corrector Design Parameters Scan frequency 13.00816 in rows   27  .

Each band is a separate array.3. The higher temperatures are backup in case the Radiative Cooler efficiency degrades.5 Relay Optics The Relay Optics consists of a graphite-epoxy structure containing a folding mirror and a spherical mirror. which are used to relay the imaged scene from the prime focal plane to the Band 5. 6 and 7 detectors on the cold focal plane.6 Cold Focal Plane The Cold Focal Plane Assembly is mounted on the cold stage of the Radiative Cooler. The Relay Optics has a magnification of 0. The Cold Focal Plane Assembly contains the detector arrays for Bands 5.5 F/# 3. and operates at a nominal temperature of 90°K~92°K and can be controlled to one of three set points (90° K. 6 and 7. The Band 6 preamplifier and the remainder of the Bands 5 and 7 preamplifier are mounted on a circular housing adjacent to the radiative cooler. The three arrays are mounted on a monolithic substrate. SiO Mirror coating overcoat 3. This magnification is used because of the reduced physical size of the band 6 detectors.5.0 Mirror material ULE glass Aluminum. The Band 6 array contains 8 detector elements. 95° K or 105° K) by a heater on the back of the substrate.3. The characteristics of the Cold Focal Plane Assembly are listed in Table 3. This photoconductive array shows a significant decrease in responsivity from 90° K to 105° K. The Band 6 array is fabricated from mercury cadmium telluride. The Band 5 and Band 7 arrays contain 16 detector elements. The input stages and the feedback resistors of the preamplifier for the Band 5 and 7 (photovoltaic) detectors are mounted with the detectors on the cold stage of the Radiative Cooler. The nominal spatial resolution of Bands 5 and 7 is the same as Bands 1 through 4.6.5 Relay Optics Design Parameters Folding mirror clear aperture outer 3.14 in diameter Folding mirror clear aperture inner 0.537 in diameter Spherical mirror clear aperture diameter 5.5. The   28  . The characteristics of the Relay Optics are listed in Table 3. The position and tilt of the spherical mirror may be varied during integration and test by three electro-mechanical linear actuators to provide focus of the Cold Focal Plane Array and to obtain the correct band-to-band registration of the cold focal plane detectors to the prime focal plane detectors.538 in Magnification . and are fabricated from indium antimonide. These photovoltaic indium antimonide arrays show very little change in responsivity from 90° K to 105° K. Table 3.3.

an intermediate stage. and the Earth. 6 and 7 preamplifiers which are inside.00190 in x 0.0 µrad Center to center spacing along 0.4 µrad 42. a radiation shield and a combination earth shield and cooler door. the sun.00408 in in Detector Area IFOV size 42. This field of view must remain free of any spacecraft structures.00204 Detector Size 0.00510 in 0. The cold stage outgas heater also serves as a backup for the cold focal plane heater.00408 in x 0. Table 3. Table 3.00408 in track Center to center spacing between 0.7 Radiative Cooler The Radiative Cooler cools the cold focal plane by radiating heat to cold space.5 µrad x 85. The flat rectangular corners of the Radiative Cooler structure that extend beyond the main circular cross section serve as radiation elements to dissipate heat from the Band 5.   29  .connections within the Radiative Cooler from the cold focal plane area to the ambient temperature connectors for the preamplifiers are made via special low thermal conductivity ribbon cables.6 Cold Focal Plane Design Parameters Parameter Bands 5 and 7 Band 6 Number of detectors 16 per band 8 0. Temperature controlled outgas heaters (controlled to 318° K) are mounted on both the cold and intermediate stages to provide temporary heating of the cold surfaces should on-orbit contamination occur.3. It has a cold stage. In this table it can be seen that the Cooler has a large field of view to cold space.00204 in 0.7 shows the cooler characteristics.01020 in rows 3.5 µrad x 39.

are as shown in Table 3.Table 3. The cold focal plane assembly has a filter housing that contains filters for band 5. located immediately in front of each detector array.3. 6. The spectral filters. 105° K 3. for a 705 km satellite altitude. The prime focal plane assembly has a filter housing that contains filters for bands 1 through 4 and the panchromatic band. and 7. are the dominant items that establish the optical bandpass for each spectral band.both 318° K stages Cold stage backup temperatures 95° K. the spectral filters. and the detector response.both temperature .8 Spectral Filters The nominal wavelength location of the ETM+ spectral bands and the nominal ETM+ IFOV size and associated ground resolution. The spectral filter for the ETM+ are of ion assisted deposition (IAD) design to reduce changes due to outgassing in a vacuum.   30  . The ETM+ spectral bandwidth is determined by the overall combination of all of the optical elements.7 Radiative Cooler Design Parameters Horizontal field of view 160° Vertical field of view 114° 660 cm Intermediate stage radiator area squared 435 cm Cold stage radiator area squared Nominal intermediate stage temperature 134° K Cold stage temperature 91° K Cold stage minimum temperature 82° K Outgas temperature .8.

005 42.5±4.3 30 5 1.005 42.005 .12.50±0.090±0.010 42.0.005 .3 30 2 0.3 1 0.515±0.005 42.605±0.3 30 6 10.630±0.5±4.350±0.690±0.900±0.020 42. and Ground Resolution Spectral Half-Amplitude Bandwidth IFOV Size Sub-satellite Band (µm) (µm) Ground Resolution (m) Panchromati 18.3 30 4 0.0 60 7 2.100 .Table 3.005 .3 30 3 0.010 13 x 15 c 21.0.0±9.020 .0.5±4.1 .525±0.100 85.775±0.0.0.40±0.Landsat 7 Project End-to-End Dataflow   31  .2.520±0.5 x 0.1.005 42. IFOV Size.010 .750±0.005 .550±0.5±4.450±0.900±0.5±4.3±4.010 .5±4.1 Ground System Overview       Figure 4.3 30   4.8 ETM+ Spectral Bands.

the Image Assessment System (IAS). The ground system context diagram illustrates both unique and institutional components and their data flow relationships end-to-end.2 . located at Goddard Space Flight Center (GSFC) in Greenbelt. Figure 4. and supervisor/managers. and resolves spacecraft anomalies. Landsat Ground Station (LGS). 4. Svalbard. Maryland is the focal point for all space vehicle operations.2 Unique Ground System Components 4. hardware. and the NASA Integrated Support Network (NISN). The unique components include the Mission Operations Center (MOC). The MOC Picture taken by Trond Løkke on March 2. and personnel required to accomplish Landsat 7 planning and scheduling. mission planners. and orbit determination are provided by the Flight Dynamics Facility. The MOC is staffed by the Flight Operations Team (FOT). software. subsystem engineers. Flight dynamics functions such as maneuver planning. the Level 1 Product Generation System (LPGS).The Landsat 7 ground system consists of both Landsat 7 unique components as well as institutional services. investigates. The Flight Dynamics Facility (FDF).   32  . 1998 also detects.Svalbard Satellite Tracking Station (Svalsat) and maintain flight software. monitor its health and status. and the international ground stations (IGS). MOC ground software. planning aid generation. The MOC provides the facilities. analyze the performance of the space vehicle. The National Centers for Environmental Prediction (NCEP). and to command and control the Landsat 7 space vehicle.2.1 Mission Operations Center The MOC. the EROS Data Center Distributed Active Archive Center (LP-DAAC). Spitsbergen Island. The institutional support systems consist of the Landsat Ground Network (LGN). Space Network (SN). which is comprised of console analysts. procedures. and located on Plateau Berger. Landsat Processing System (LPS).

into separate accumulations of Earth image data.2 Landsat Ground Stations The LGS. South Dakota is a receive site for the wideband XBand downlinks of payload data from the space vehicle.In addition to the ground sites. once for each format. and the detector offsets inherent to the focal plane array engineering design. One channel holds bands 1 through 6. South Dakota records all wideband data. the Alaska Ground Station (AGS) and the Svalbard Norway Ground Station (SGS) receive payload data downlinks. 4. the pixel shift parameters used are documented in the IAS CPF. The LGS also supports S-Band command telemetry operations. in Sioux Falls. respectively. The LPS spatially reformats Earth imagery and calibration data into Level 0R data. the Tracking Data and Relay Satellites (TDRS). which serves as a front-end processor. Channel accumulations represented by bands 1 through 6 and 6 through 8 become formats 1 and 2. tracking services and spacecraft clock maintenance capabilities are provided. PCD and MSCD are generated twice. the other LGN components are institutional services. In addition. Payload data downlinked to SGS and AGS is captured on tape and shipped to LGS.2. All LPS 0R corrections are reversible. into its wideband data stores. This involves shifting pixels by integer amounts to account for the alternating forward-reverse scanning pattern of the ETM+ sensor. calibration data.   33  . in Sioux Falls. each at a different frequency. and payload correction data (PCD). An I-Q channel pair represents a complete data set. and transmits the acquired wideband data over four 75-Mbps LGS output channels to the LPS where they are recorded. Their contents should be identical. mirror scan correction data (MSCD). The LGS is a Landsat 7 unique component of the Landsat Ground Network (LGN). at realtime rates. LGS separates each X-band data into two 75-Mbps channels ( I and Q). The LPS retrieves and processes each channel of raw wideband data. and the other holds bands 7 and 8 and a second gain form of band 6.3 Landsat Processing System The LPS located at EDC. located at EDC. 4. as well as tracking.2. In addition to LGS. Together these sites provide the ability to downlink real-time and stored housekeeping data (S-band). The NASCOM division of NASA provides all network lines. operated by the NASA's Space Network are utilized. and to command the spacecraft. The LGS acquires ETM+ wideband data directly from the Landsat 7 spacecraft by way of two 150 megabit-per-second (Mbps) X-band return links. the odd-even detector arrangement within each band. at lower than real-time rates.

In addition to its assessment functions.5 Image Assessment System The IAS. Enhanced geometric accuracy is possible with the application of ground control and terrain models but these are not used by LPGS.4 Level 1 Product Generation System The LPGS. and output format. A number of user-selectable options exist for configuring a 1G product. generates Level 1 products in response to user requests received from the LP-DAAC. calibration data. PCD. located at EDC. and metadata are structured into HDF-EOS for each format and sent to the LP-DAAC for long term archival in subinterval form. These details can be found in the chapter on Data Products. resampling methodology. Users can order either 1R (radiometrically corrected only) or 1G (radiometrically and geometrically corrected) products. product size. MSCD. Radiometric and geometric processing is performed by LPGS on Level 0R data to create Level 1 products.2.2. The IAS periodically performs radiometric and geometric calibration and updates the CPF. radiometrically corrected. located at EDC. The two formats of data are united when a Landsat 7 0R product is ordered. Metadata are generated for the entire subinterval and on a scene-by-scene basis. The CPF also is sent to international ground stations via the MOC. map projection. rotation.During LPS processing. Operational activities occur at EDC while less frequent assessments and calibration certification are the responsibility of the Landsat-7 Project Science Office at the   34  . grid cell size. 4. The image data. This file is stamped with applicability dates and sent to the LP-DAAC (EDC) for storage and eventual bundling with outbound Level 0R products. is responsible for the off-line assessment of image quality to ensure compliance with the radiometric and geometric requirements of the spacecraft and ETM+ sensor throughout the life of the Landsat-7 mission. format 1 bands are duplicated. and used to assess cloud cover content and to generate a browse image. Cloud cover scores are generated on a scene-by-scene and quadrant-by-quadrant basis. The browse image is sent to the LP-DAAC separately for use as an online aid to ordering. These include band selection. The CPF supplies the radiometric and geometric correction parameters required during Level 1 processing to create superior products of uniform consistency across the Landsat 7 system. 4. The 1G products are classified as systematic meaning the class of corrections applied is derived from spacecraft data only. the IAS is responsible for the radiometric and geometric calibration of the Landsat 7 satellite and ETM+.

the IGSs return metadata for their station holdings to the LP-DAAC. South Dakota is part of the Earth Observing System (EOS) Data and Information System (EOSDIS).7 Land Processes Distributed Active Archive Center The LP-DAAC. processing. and data archival and distribution functions for a variety of data types including Landsat 7.6 International Ground Stations   The IGSs are satellite data receiving stations located around the world. Although catalogued at EDC. user interface.2. In addition to the Level 0R data received from LPS. 4. The IGSs submit downlink requests to the MOC and receive schedule and orbital element data from the MOC. the LP-DAAC also receives calibration parameter files from the IAS. The acquisition circles for the IGSs depict the Earth's land areas that are regularly imaged. and distribution services for their user community. Maryland.Goddard Space Flight Center (GSFC) in Greenbelt. located at Sioux Falls.3 .   35  . They provide data receive.2. They receive Landsat 7 payload data via X-Band direct downlink. In addition. The X-Band direct downlink data includes the PCD required for image processing. Figure 4. data downlinked to the IGSs must be order from these foreign stations. It provides information management.IGS Circles   4. The LP-DAAC performs billing and accounting functions and serves as the repository for user-oriented documentation.

provides space-to-space and space-to-ground data relay services. is the starting point for nearly all weather forecasts in the United States. and SN tracking data preprocessing.   36  . SN. and SGS. 4.2 Space Network The SN. and generation of planning and scheduling aids (including in-view predictions for IGSs.3.3. For Landsat 7. which provide S-Band and X-Band communication support to the Landsat 7 mission.4. WPS. the Wallops Island.3 National Centers for Environmental Prediction The National Centers for Environmental Prediction (NCEP) provide timely. 4.3 Institutional Ground System Components 4. Virginia (WPS) ground station. a critical part of the National Oceanic and Atmospheric Administration’s National Weather Service. These are used for Landsat 7 real-time command and telemetry monitoring during on-orbit operations on a scheduled basis and possible emergency operations on a callup basis. provides workstations in the MOC which are used by the Flight Operations Team (FOT) for orbit determination. maneuver planning support.1 Landsat Ground Network LGN consists of multiple communications sites. the NCEP supplies cloud cover predicts data to the MOC for image scheduling. an institutional support element located at GSFC. ephemeris data generation. The LGN institutional services include AGS. The SN is managed by GSFC. local oscillator frequency reporting. NCEP generates weather related products.4 Flight Dynamics Facility FDF. attitude determination. accurate and continually improving worldwide forecast guidance products. 4. NCEP. and the SGS in Norway.3. The SN collects Landsat 7 space vehicle tracking data for FDF processing. The FDF institutional facility retains responsibility for star catalog maintenance. The Wallops Facility manages the AGS.3. which includes the Tracking and Data Relay Satellites (TDRSs) and the ground terminals at the White Sands Complex. and the LGN).

The single integrated network project is replacing the independent special purpose networks that have served individual customers for decades. The spacecraft crosses the Equator from north to south on a descending orbital node from between 10:00 AM and 10:15 AM on each pass.3. 5. This same day orbit configuration will space the satellites ideally 15 minutes apart (i. A multispectral data set having both high (30 meter) and medium to coarse (250 to 1000 meter) spatial resolution will thus be acquired on a global basis repetitively and under nearly identical atmospheric and plant physiological conditions.5 NASA Integrated Support Network NISN is a global system of communications transmission switching and terminal facilities that provide NASA with long-haul communications services.5 km/sec. Landsat 7 and Terra were launched and injected into identical 705 kilometer.1 . voice. The spacecraft completes just over 14 orbits per day. and video information in the most cost effective manner possible. and near polar at a nominal altitude of 705 km (438 miles) at the Equator. each orbit takes nearly 99 minutes. circular.1 Orbit Figure 5. covering the entire Earth between 81 degrees north and south latitude every 16 days.1 illustrates Landsat's orbit characteristics. NISN was implemented to serve the needs of all of NASA's users for the transmission of digital data. The NISN supports the above institutional facilities. sunsynchronous orbits in 1999. equatorial crossing times of 10:00 to 10:15 AM for Landsat 7 and 10:30 for Terra).4. Circling the Earth at 7.   37  . Figure 5.Landsat Orbit The orbit of Landsat 7 is repetitive. Sun-synchronous.e.

The 16-day Earth coverage cycle for Landsat 7 is known as the swathing pattern of the satellite Figure 5. Moving from the Equator toward either pole. The ETM+ sensor onboard the spacecraft obtains data along the ground track at a fixed width or swath as depicted in Figure 5. it becomes straightforward to select Landsat 7 scenes or subintervals required for a specific project.3 83.1 Image Sidelap of Adjacent Swaths Latitude (degrees) 0 10 20 30 40 50 60 70 80 Image Sidelap (%) 7.ETM+ Swath Pattern!   38  .9 19. adjacent swaths overlap at the edges by 7.3 .    Figure 5.3.3 percent.1.ETM+ Swath Figure 5.9 At the Equator.2 .3 8.2 Swathing Pattern Landsat 7 orbits the Earth in a preplanned ground track as illustrated in Figure 5.4 53.2. As seen in this figure for daytime acquisitions. Landsat 7 travels the adjacent swath to the west of a previous swath one week later (and the adjacent swath to the east occurred one week earlier and will recur nine days later). Table 5.7 29. this sidelap increases because the fixed 185 km swath width.0 40. Table 5.6 68.1 shows the amount of sidelap from 0 to 80 degrees latitude in 10-degree increments.7 12.5. After familiarization with the data acquisition cycle or swathing pattern.

Figure 5. Row one of each path starts at 80° 47' N and the numbering increases southward to latitude 81° 51' S (row 122). the continuous data stream or subinterval is framed into individual scenes each 23. As the spacecraft moves along a path. This includes the ascending rows. the row numbers ascend northward. The rows have been assigned in such a way the row 60 coincides with the Equator (descending node).5.4 . The WRScornerPoints.4 graphic depicts the Landsat path/row schema. During ground processing. Then. The term row refers to the latitudinal centerline across a frame of imagery along any given path. The WRS indexes orbits (paths) and scene centers (rows) into a global grid system (daytime and night time) comprising 233 paths by 248 rows. beginning with row 123.   39  .3 The Worldwide Reference System The standard worldwide reference system as defined for Landsat 4 and 5 was preserved for Landsat 7. cross the Equator (row 184) and continue to latitude 81° 51' N (row 246). It is sorted in Path/Row order. Row 248 is located at latitude 81° 22' N. negative longitude is West.92 seconds of spacecraft to create 248 rows per complete orbit.xls file lists the latitude and longitude for the scene center and four corners of each WRS scene. the ETM+ scans the terrain below. Positive latitude is North. Positive longitude is East.WRS Path/Row Numbering Scheme Successive orbits and spacecraft attitude are controlled to assure minimal variation to either side from the intended ground track and framing of scene centers is controlled through LPS processing so that successive images of a specific scene or scenes can be registered for comparison purposes. negative latitude is South. Figure 5. The Landsat satellites are not placed in a true polar orbit but rather a near polar orbit which means the path/row numbers do not coincide with latitudes 90° north and south. whereupon the next path begins.

The WRS-2 system divides the Earth into a grid of 57.5-month chunks. Instead. By applying a set of algorithms on a daily basis. The Landsat-7 observatory is operated such that it follows the WRS grid within tight tolerances. the LTAP is designed to ensure optimal collection of Landsat-7 Enhanced Thematic Mapper-Plus (ETM+) imagery for scientific applications. and Arvidson et al.4 Long Term Acquisition Plan 5.EROS Data Center Sioux Falls.WRS path/row maps are available from EROS Data Center and have the LANDSATs 1-3 WRS on one side and the LANDSATs 4-5-7 WRS on the reverse. 1999 and has been extended out to February 22.cr.usgs. A second chart.2 LTAP Overview The ETM+ does not acquire data continually. Gasch and Campana (2000). SD 57198 Tel: 605-594-6151 (7:30 am to 4:00 pm CT) TDD: 605-594-6933 (7:30 am to 4:00 pm CT) Fax: 605-594-6589 (24 hours) custserv@edcmail.4.   40  .784 scenes.4. Contact EDC Customer Services to request WRS maps: Customer Services USGS . substantially cloud-free land images. should consult Goward et al. (1999). 5. 2004. This section provides a brief overview of the LTAP and details the specific algorithms and input files used for calculating scene acquisition priorities. The map sheets are at 1:10 million scale and 26 are needed for global coverage. while minimizing the effects of cloud-cover and system constraints. It starts on May 27. 5. dated January 13. The latest revision number for these charts is 7. with larger type. The user to extend further in the future easily updates the charts.gov An Excel spreadsheet allows users to look up the calendar date on which a particular path will be followed by Landsat 7 (listed in blue) or by Landsat 5 (listed in red). 2006. (2001). including the science justification for the approach and recent performance results. acquisitions are scheduled in advanced using a Long Term Acquisition Plan (LTAP) in conjunction with a software scheduler. with each scene centered on the intersection of a path (groundtrack parallel) and row (latitude parallel).1 Introduction The Landsat-7 Long Term Acquisition Plan (LTAP) automates the selection of Landsat scenes to periodically refresh a global archive of sunlit. Those wishing a more complete review of the LTAP. is also available in this file that contains the path flow/calendar dates in 3.

S. Scene clustering 7. nominal cloud-cover 3. and all known islands and reefs.000 scenes containing "land" was compiled for the LTAP. Missed opportunities for previous acquisitions 5. urgency of acquisition. recorder capacity. Factor 4 requires feedback between the list of archived scenes and the LTAP. while factor 5 requires feedback between cloud cover assessment of archived scenes and the LTAP. such as growth and senescence of vegetation.   41  . while the actual acquisition list is obtained by incorporating system constraints (factor 7). The Landsat-7 LTAP includes several aspects that influence whether a particular scene should be acquired or not. such as when full growth canopy exists or during winter quiescence. System constraints (duty cycle.3 Seasonality An objective of the LTAP is to schedule acquisitions more frequently during periods of change. Seasonality of vegetation. when only a single acquisition is necessary. Each of these factors is discussed in the following sections. A database of just over 14. niche-science communities 2. etc).. 1994. archive to 250 scenes. To flag seasonal change within a WRS scene. when multiple acquisitions are meaningful. Quality (cloud-cover) of previous acquisitions 6. Predicted vs. These include: 1. An 8year AVHRR Normalized Difference Vegetation Index (NDVI) data set was used to determine where and when change was occurring (Goward. a statistical test was applied to each NDVI 45-km sample mapped to that scene. Mission resource limitations (discussed later) restrict the daily acquisition volume for the U. 1999). approximately 850 of these WRS land scenes (descending. Comparing each sample or pixel to the same pixel two months later revealed periods of significant change in the NDVI mean and standard deviation. and availability of resources to optimize fulfillment of the Landsat mission goals. Sun angle 4. and periods of minimal change. 5.overflying the entire global WRS scene grid every 16 days. basing these decisions on cloud-cover forecasts.4. Antarctic sea-ice. including continental areas. The mission scheduler must select the "best" 250 scenes for acquisition each day within these constraints. et al. ground station locations. and less frequently during relatively stable periods. The scheduler automatically schedules acquisitions in accordance with the LTAP. shallow coastal waters. Within any given 24-hour period. Factors 1-6 feed directly to calculating the priority of each potential acquisition. sunlit paths only) are in view of the ETM+ and are candidates for acquisition.

and 3% are devoted to night images or special high-priority acquisitions (natural disasters. approximately 27% of the 250 scenes/day acquired by Landsat-7 are devoted to satisfying "niche" requests.4 Predicted Cloud-cover The cloud-avoidance strategy is based on the relative predicted cloud cover with respect to the seasonal average cloud cover for each scene. Requirements for night imaging and high-priority acquisitions are dealt with elsewhere in the data specification.Thus. that path-row location is no longer eligible for acquisition. etc). These locations currently include: • • • • • • • • • • • 282 agriculture (acquire every season if CC predict < 60%) 35 calibration (acquire "always") 896 reefs (acquire from 2x to 6x each year) 30 fire (acquire "always") 1392 land ice (acquire once during certain months) 3601 Antarctica (acquire once during Jan-Feb) 60 oceanic islands (acquire twice each year) 1175 rainforest (acquire "always" all year) 352 sea ice (acquire from 1x to 3x each year) 11 Siberia (acquire "always" over 9 months) 72 volcanoes (acquire from 2x to 12x during year. Rather. that scene will always be a candidate for acquisition. "acquire always". Niche Science Communities. The LTAP scheduling software compares near-term predictions to the historical average cloud cover for that scene for that month. national needs. system constraints. defining acquisitions solely through semi-monthly NDVI change does not fully capture the science and user interest in the Landsat mission. and other factors within the LTAP (cloud-cover. etc) will govern whether or not an acquisition actually takes place. approximately 70% are devoted to satisfying routine acquisitions from the seasonality file. a location may be labeled as "acquire once". specific niche science communities have requested locations for more frequent acquisitions. "acquire once. night) The exact makeup of these niche science acquisitions varies each quarter of operations.4. To supplement the NDVI-based seasonality metric. In general. depending on user needs and science priorities. Scenes with better-than-average cloud-cover   42  . 5. the year is broken into a set of temporal "windows" for each path-row location. During each window. to determine acquisition priorities each day. Conversely. or "never acquire".” means that once a successful acquisition occurs during the time window. incl. It should be noted that "acquire always" does not imply that an actual acquisition will occur for every overpass during that time window. and this information is stored in the LTAP seasonality file.

5 Sun Angle Requests for high-latitude scenes are rejected during local winter when the solar elevation is below a threshold. Figure 5. 1994).6 Missed Opportunities A request is granted a priority boost as a function of the number of consecutive past cycles in which the opportunity to acquire the requested scene was not fulfilled. when an acquisition window opens) are given a priority boost. that give monthly estimates of cloud-cover percentage for 2.4. This threshold was set at 5° during the initial years of the survey.7° grid scale (Campana.514 unique path/row combinations !in the U. 1996). The forecast nearest in time to a candidate acquisition is compared to the ISCCP climatology to determine whether the predicted observing conditions are better or worse than typical for that location..predictions are given a priority boost. For example. 2000).   43  .    5.July 11. then a request for this scene is granted a priority boost based on two missed opportunities from 32 and 16 days ago. The cutoff angle remains at 5° for the southern hemisphere. if the last successfully acquired image of a scene was 48 days ago. these cloud predictions are translated to the Landsat WRS coordinate system. The cutoff angle was changed to 15° for the northern hemisphere on July 24. In addition. Acquisition priority is then adjusted upward to favor the better conditions or downward to avoid the worst conditions. 2002 because of duty cycle concerns and the fact that snow dominates scenes acquired at lower angles.This map shows the lowest available cloud!cover for each of the 16. Global historical cloud data were obtained from ISCCP records (Rossow et al. or when the acquired image fails to meet minimum quality standards. while those with poorer cloud cover predictions are given a lower priority.5 illustrates the cloud-free nature of the first year ETM+ archive. . such as cloud cover. all new requests (i. The forecasts are compiled at 0600 UT each day and apply to each 3hour interval up to 84 hours. 5.e.4. Once received from NOAA. This occurs when the scene was not scheduled for acquisition. Monthly averages over the 5 years were computed and mapped to the Landsat WRS grid to produce a cloud climatology (Figure 5).5° grid cells from 1989 to 1993. The NOAA National Centers for Environmental Prediction provide daily global forecasts of cloud-cover percentage at a 0. Figure 5.S. Landsat 7 archive (June 28.5 .

EROS Data Center. 3. and the frequency of acquisition during those periods. as determined during image processing.4.0.8. 4.7 Image Quality Quality (cloud-cover) of the best or most recently acquired image of a path-row from past cycles factors into the relative demand for another acquisition. A more recent archive scene count (as of January 31. Quality is a function of cloud contamination in the image.S. 2003) is presented in Figure 5.5.4. The lowest cloud cover scores for each WRS location are plotted in Figure 5.8 Clustering Within the Landsat-7 LTAP.6). 5. This (i) reduces the on-off cycling of the ETM+ instrument and (ii) promotes the archiving of continuous swath data. As of July 31. at this server location. This file covers the year 2001 and includes both daytime and nighttime imaging. The seasonality file specifies which WRS scenes are to be acquired during which periods of time (request period). 2. the scheduler also considers the quality of the best image that could be spliced together as a mosaic of imagery of adjacent scenes acquired in the recent past. The results of ingesting this file into the scheduler and developing the daily schedule are posted daily on the web. 6.9 Long Term Acquisition Plan Files The Long Term Acquisition Plan (LTAP) is used to direct the acquisition by Landsat 7 of Worldwide Reference System (WRS) scenes around the world for archiving in the U. 2002 the number of scenes successfully archived was 252. 5. 3. Frequency of acquisition is defined as either "once" during the request period.7. or "all" opportunities during the request period. October.   44  . 2002 nominal cloud cover file nominal cloud cover daylight additions file nominal gain settings file maximum solar zenith angle 1.4. a higher priority is given to scenes that form "clusters" contiguous groups of along-path acquisitions. which in turn allows "floating scene" subsetting of user-defined areas without regard for scene boundaries. 5. The contents of the following files represent the underpinnings of the LTAP: 1.208 (Figure 5. Rev. seasonality file WRS land data base. At high latitudes.

Here is the warning label associated with the seasonality file: The scheduler has many resources and priorities that it juggles during consideration of requests for scheduling. onboard recorder space. should the first acquisition's cloud cover be considerably worse than the nominal. As subsequent years are processed. The average cloud cover is derived from the International Satellite Cloud Climatology Project D2 data set. political (disputed sites). including: • • • • • predicted cloud cover as compared to the nominal number of missed opportunities for this request since it was opened how good the cloud cover assessment score was for the last acquisition nearness to end of the request period availability of resources including duty cycle. glaciers.   45  . The nominal cloud cover daylight additions file adds those ascending rows. and station contact time The result of this juggling is that scenes marked with an "all" opportunities frequency are usually acquired every 4-5 cycles (64-80 days) instead of every cycle (16 days). MODIS fire validation sites. global oceanic islands. 3. which will be in daylight during some part of the year and therefore may be imaged. The file spans one year and addresses the descending rows only. which will be imaged at least once every year. volcanoes. all) base priority of the request maximum solar zenith angle (only for night scenes. this file will be updated. years 19891993. At the end of the file are separate tables for the niche communities and special interests. These include: EOS validation sites. So please treat the frequency assignment in the seasonality as a guide. A comprehensive reef list is a future addition. and 4. is set to 90) 2. The nominal cloud cover file reports the average cloud cover for each WRS scene for each month of the year. The WRS land database identifies those WRS scenes containing land or shallow water. not a rule. Another outcome of this is that scenes marked with a "once" frequency may be acquired multiple times within the request period in an attempt to better the cloud cover results. The column headings for the seasonality file are: • • • • • • first date of request period (yyyy-mm-dd) last date of request period (yyyy-mm-dd) path/row (note that rows 246-248 will be sorted to the end of the path) acquisition frequency (once.

(Powerpoint format) 6. five types of interaction are possible. 5. Transmitted . once for format 1 and once for format 2. The file addresses both descending and ascending rows of interest to the U. 3.S.(ASCII text) • monthly gain maps . The default gain settings are generated using rules that take land cover type and sun angle into account. The maximum solar zenith angle is a single value for all scenes at all times of the year.The energy is returned unchanged with the angle of incidence equal to the angle of reflection. Daylight imaging will not be scheduled if the angle is 85 degrees or more (same as 5 degrees or less in elevation from the earth's surface). 6.(ASCII text) • arctic mask . Other files referenced by this document are: • desert mask . 2. To understand these rules.The energy is given up to the object through electron or molecular reactions. You will notice that bands 66 are constant at LH and band 8 is constant at L. The order in which they are specified for each scene is bands 123456678.1 Scientific Theory of Measurements When solar energy strikes an object. Reflected . The settings are given as H for high gain or L for low gain. with bands 123 changed together. When we do night imaging. The nominal gain settings file identifies the gain settings that will be used as defaults for each WRS scene for each day of the year. Reflectance is the ratio of reflected energy to that incident on a body. archive. where band 6 is repeated twice.The energy passes through with a change in velocity as determined by the index of refraction for the two media in question.The columns in the nominal cloud cover files are: first day of month|last day of month|path|row|cloud cover value all cloud cover values for row 1 across all paths are given. The DD-MMM values that are included in the file are for a leap year and are guidance for those who don't easily equate a DOY value of 177 with the date 25-JUN. and bands 57 changed together. The wavelength reflected (not absorbed) determines the   46  . The energy is: 1. we use the following default settings: HHHHLLHLL. read the file "gain setting rules". This file spans one year. then for row 2. The file only contains entries showing when the default gain setting has changed from its previous setting. etc. Absorbed . The other bands are changed in groupings. band 4 changed on its own merits. The scheduler uses the DOY value. and band 8 is the panchromatic band.

Emitted . the energy is first absorbed. A signature on such a graph can be defined as reflectance as a function of wavelength. reflectance.1 . Radiance corresponds to brightness in a given direction toward the sensor. usually at longer wavelengths. The Landsat-7 system is designed to collect 7 bands or channels of reflected energy and one channel of emitted energy. Rayleigh and Mie scatter are the two most important types of scatter in the atmosphere. according to the basic premise. A plot of the collective interactive mechanisms (scattering.Spectral Reflectance Curves of Four Different Targets   47  . The basic premise of using spectral signatures is that similar objects or classes of objects will have similar interactive properties with electromagnetic radiation at any given wavelength. Radiance refers to the flux of energy (primarily irradiant or incident energy) per solid angle leaving a unit surface area in a given direction.color of an object.The direction of energy propagation is randomly changed. A well-calibrated ETM+ enables one to convert the raw solar energy collected by the sensor to absolute units of radiance. 4.Actually. Scattered . which is the ratio of reflected versus total power energy. the amount of emitted and reflected radiation varies by wavelength. emittance.1. and then re-emitted. The object heats up. The eight bands of ETM+ data are used to discriminate between Earth surface materials through the development of spectral signatures. and is often confused with reflectance. For any given material. These variations are used to establish the signature reflectance fingerprint for that material. or spectral signature. different objects will have different interactive properties. Radiance is what is measured at the sensor and is somewhat dependent on reflectance.     Figure 6. Conversely. Four such signatures are illustrated in Figure 6. result in a unique curve. that is diagnostic of the object or class of objects. 5. and absorption) at wavelengths on the electromagnetic spectrum should.

focal length.5 µm.2 using just two of the ETM+ spectral bands. and SW representing silty water) may be characterized as distinct.45 to 12. Figure 6.ETM+ data can be used to plot spectral signatures although the data are limited to eight data points within the spectral range of . (GL representing grasslands. ETM+ has an IFOV of 30   48  . 30 meters for bands 1-5. More commonly used descriptive terms for spatial resolution are ground sample distance (GSD) and instantaneous field of view (IFOV). When more than two wavelengths are involved.2 Spatial Characteristics Spatial resolution is the resolving power of an instrument needed for the discrimination of features and is based on detector size. is the area of terrain or ocean covered by the field of view of a single detector. and sensor altitude.Spectral Separability Using Just Two Bands  Each of the materials has been plotted according to its percent reflectance for two of the wavelengths or spectral bands. or pixel size. the plots in multi-dimensional space tend to increase the separability among different materials. IKONOS. 6. More useful is plotting the ETM+ spectral signatures in multi-dimensional feature space. RS representing red sand.2 .3 illustrates the ETM+ IFOV for bands 1-5 and 7 relative to other sensors and a football field. has an IFOV of 1 meter. the recently launched Space Imaging sensor. This spectral separation forms the basis for multispectral analysis where the goal is to define the bounds of accurately identified data point clusters. and 15 meters for band 8.   Figure 6. The four surface materials shown in Figure 6. PW representing pinewoods. The French SPOT panchromatic sensor an IFOV of 10 meters whereas the SPOT multispectral (XS) sensor has an IFOV of 20 meters. The IFOV.1 are plotted in Figure 6. and 7. 60 meters for band 6. The ETM+ ground samples at three different resolutions.

000 8 It is natural to assume that one could determine a scene's spatial extent by multiplying the rows and columns of a scene by the IFOV.meters for bands 1-5.1 Table 6.600 6000 8 6 60 3. While this calculation applies to scene length. Figure 6.25 meters. The staggered image bands result from the focal plane design. which LPS accounts for by registering the bands during 0R processing.1 Image Dimensions for a Landsat 7 0R Product Sample Band Resoluti Data s Bits per Numb on Lines (column Sample er (meters) (rows) s) 1-5.200 12. The end result is an increasing amount of zero-fill preamble according to the band order on the ground projected focal plane array. Left and right image buffers were placed in the 0R product to accommodate a possible increase in scan line length over the mission's life.ETM+ Spatial Resolution Relative to Other Sensors A standard WRS scene covers a land area approximately 185 kilometers (across-track) by 180 kilometers (along-track). the scene width calculation is more complicated due to the presence of image buffers and the staggered image bands in the 0R product.3 . 7 30 6.300 3. These are listed in table 6.000 8 8 15 13. A more precise estimate for actual scene size can be calculated from the 0R product image dimensions. This would lead to a scene width of 198 kilometers (6600 samples * 30 meters) and a scene length of 180 kilometers (6000 lines * 30 meters). and 7 of 30 meters while the Indian Remote Sensing Satellite (IRS) has an IFOV of 36.   49  .

This number is determined by looking at the reverse even detector offset for band 8. Table 6.0 124.0 194.0 126.0   50  .0 98.0 110. The number of coincident images pixels in a scan is therefore 6087 (6333 .0 113.0 99. Coincident imagery for all 8 bands starts at pixel location 247 for the 30-meter bands.666.0 114.0 76. is halved to put the ending pixel number into 30-meter units.247 + 1). 12.0 74.0 101.The detector offsets determine the amount of zero fill preamble for each band.0 50. Coincident imagery for all 8 bands ends at pixel location 6333 for the 30-meter bands.0 116. which translates to 232 30-meter pixels.0 120. 116.0 Band Numb er 1 2 3 4 5 6 7 8 Reverse Scan Odd Detectors 48.626 that represents the number of band 8 pixels per line (6313 minor frames times 2).0 195.0 171. These are listed in Table 6.0 197.2 and can also be found in the Calibration Parameter File.61 kilometers (6087 * 30 meters). is actually in 60 meter IFOVs. The total. This number.0 44.0 165.0 45.2 ETM+ Detector Shifts Forward Forward Scan Reverse Scan Scan Even Even Odd Detectors Detectors Detectors 49.0 73. Add to this number the value 12.0 54.0 70.0 51.0 123.0 95.0 40. Another 14 pixels must be added to this number to account for the seven minor frames of image data pre-empted by time code. The nominal width for a scene is therefore 182.0 168.0 191. One need only to look at the reverse scan odd detector offset for band 6 to see that this is true.0 169.

the mean sun time at each individual point in the orbit will remain fixed. as defined in figure 6. The actual effects of variations in sun elevation angle on a given scene are very dependent on the scene area itself. The local time that the satellite crosses over a given point at latitudes other than at the equator also varies due to the time the satellite takes to reach the given point (nearly 99 minutes are required for one complete orbit). As a result.6. The   Figure 6.4 .5 . and the time zones crossed by the satellite relative to its equatorial crossing point.4. the nominal mean sun time of the descending node at the Equator is 10:00 AM. and in fact. These changes are due primarily to the north-south seasonal position of the sun relative to the Earth (Figure 6. all points at a given latitude on descending passes will have the same mean sun time.5.Sun Synchronous Orbit Landsat 7 A fixed mean sun time does not mean that local clock time will remain fixed for all points at a given latitude.6). Consequently. 6. Figure 6. since discrete time zones are used to determine local time throughout the world. cause variations in the illumination conditions under which imagery is obtained. track and the mean projection of the sun onto the equatorial plane will remain nearly constant throughout the mission. as shown in figure 6.2 Sun Elevation Effects While the orbit of Landsat 7 allows the spacecraft to pass over the same point on the Earth at essentially the same local time every 16 days. changes in sun elevation angle. the geometric relationship between the orbit's descending.1 Orbit Times The Landsat 7 orbit is sun synchronous.3.3.3 Temporal Characteristics 6. or southbound.Sun Elevation Angle 51  . For Landsat 7.

Apart from the variability of scene effects. Deserts do not qualify and thus are imaged once or twice per year. and adherence to the Long Term Acquisition Plan make this impossible.6 . Duty cycle constraints. sun elevation angle is itself affected by a number of perturbing forces on the Landsat orbit.   52  . it may or may not be possible to obtain useful imagery at lower sun elevation angles. Temperate forests and agricultural regions qualify as dynamic and are imaged more frequently. however. and this results in changes to the nominal sun elevation angle. and these too can vary with time of year. Figure 6. This does not mean every scene is collected every 16 days.3 Revisit Opportunities Repeat imaging opportunities for a given scene occur every 16 days (see Chapter 5 for details). has Figure 6. Depending on the scene area. These include forces such as atmospheric drag and the sun's gravity. can be considered minor for most applications. below which no imagery is acquired. is to collect as much imagery as possible over dynamically changing landscapes. The effects of orbit perturbations. is significantly more sensitive to variations in sun elevation angle than most types of vegetation. limited onboard recorder storage. one should expect that all image data can be fully exploited. The goal. At sun elevation angles greater than 30 degrees. A sun elevation angle of 15 degrees.Landsat 7 data archived during the mission's first 112 days. Although the during the first !112 days of operation. however.7 . 6. for example.reflectance of sand.3.Effects of Seasonal Changes on Solar Elevation Angle been established for the Landsat 7 mission. each general type of scene area must be evaluated individually to determine the range of sun elevation angles over which useful imagery can be realized. the use of cloud cover predictions.7 illustrates archived imagery Figure 6. Because of such factors. Atmospheric effects also affect the amount of radiant energy reaching the Landsat sensor. They have the effect of shifting the time of descending node throughout the year.

S including Alaska is quite green because every imaging opportunity is exploited. 6. A multi-temporal analysis using images such as these allows one to resolve. The importance of imaging dynamically changing landscapes frequently is illustrated in Figure 6. Northern Asia is mostly red and yellow due to recorder constraints. 1999 while the other was acquired four cycles later on October 17. This occurs in the preceding scene.4.Design ETM+ Reflective Band High and dynamic range. Low gain Dynamic Ranges therefore.9 . For all bands. The band combination for both images is 5-4-2. 1999. with greater accuracy.4 Radiometric Characteristics 6. Gain selection for a scene is controlled by the MOC and is performed by changing the reference voltage for the analog to digital convertor. The science goal in switching gain states is to maximize the instrument's 8-bit radiometric resolution without saturating the detectors. and phenological growth patterns. the low gain dynamic range is approximately 1. Table 6.   53  .mission is still young.8 . 1999 (left) and October changes in the mountains to the east of 17. to image in low gain mode when surface brightness is high and in high gain mode when surface brightness is lower.9). species components.1 Gain States The ETM+ images are acquired in either a low or high gain state (Figure 6. The dramatic color Figure 6. key landscape components such as biomass.August 14. This requires matching the gain state for a given scene to the expected brightness conditions. North Africa is mostly desert and appears red.3 lists the target/specification minimum saturation levels for all bands in both the low and high gain states. The U. !1999 (right) images of the Salt Lake City area Salt Lake City indicate the montaine growing season is over.5 times the high gain Figure 6.8. The image on the left was acquired over Salt Lake City on August 14. certain trends are emerging. It makes sense.

2 Gain Settings The currently implemented gain setting strategy for the Landsat-7 Long Term Acquisition Plan (LTAP) took effect July 13.3 193.10 235. Sea Ice 6.21 12. Within the Long Term Acquisition Plan. 1999 operational decision to acquire band 4 in low gain when the sun elevation angle exceeds 45 degrees.3 ETM+ Dynamic Range watts / (meter squared * str * µm) Low Gain High Gain Minimum Minimum Saturation Level Saturation Level 285.7 190.7 11. non-ice) 2. The earth surface categories are as follows: 1.5 17.0 291. Desert 3. and gain setting rules that are surface cover and sun angle based. Volcano/Night   54  .6 225. It consists of a fixed categorization of the surface cover types of the Earth. Land (non-desert. Ice/Snow 4.3 31. 2000 (2000/195:22:40z) superseding the at-launch plan and the December 2.0 149.3 6.0 149.4. Water 5.Ba nd 1 2 3 4 5 6 7 8 Table 6. the gains for bands 1-3 are currently always changed together as are the gains for bands 5 and 7.7 225.78 16.0 156.6 47.

65 in band 3.66 ) saturation. Band 4 set to high gain except where sun elevation is greater than 31° . Bands 1-3 set to high gain b. >0. Band 8 set to low gain 3.to avoid dense vegetation (reflectance > 0. Band 5. Band 5.Each Landsat-7 WRS Path/Row location is categorized into one of these six types. Ice/Snow and Sea Ice: a. For each surface cover type the gain setting rules are different: 1. Desert: a.7 set to high gain except where sun elevation is greater than 38° -. 7 set to high gain d. Band 4 set to high gain except where sun elevation is greater than 45°(set to low gain) . non-ice): a.03 in band 2]) saturation.to avoid bright desert target (reflectance >0. [>0.) c.to avoid bright desert target (reflectance >0. Band 4 set to high gain except where sun elevation is greater than 45°(set to low gain) . (At this sun angle high gain in band 4 saturates at about a reflectance of 0.94 in band 1.95 in band 3. Band 8 set to low gain   55  .92) saturation.68 in band 7] saturation d.71 in band 2]) saturation b.66 ) saturation. Band 8 set to low gain 2. [>0.to avoid bright desert (reflectance > 0. Bands 1-3 set to high gain except where sun elevation is greater than 28° . b.to avoid snow/ice (reflectance >0. Bands 5. >1.66. Land (non-desert. so switching to low gain keeps targets at this reflectance or below from saturating. [ >0. Bands 1-3 set to high gain except where sun elevation is greater than 19° . 7 set to high gain d.70 in band 5.to avoid snow ice (reflectance > 0. c. c.66 in band 1 .

Bands 1-4. 7 set to low gain to reduce saturation of volcanic hot spots c.. gain changes are made during flywheel periods prior to scene requiring the change and during the lowest priority scene in that period. On April 20. and shortwave infrared (5 and 7) gains were considered separately. The current implementation does not incorporate this variability.4.5% over the year producing a ±3% irradiance variation. at the original 15 km spatial resolution. Also.4. al. near infrared (band 4). no single scene gets a gain change two acquisitions in a row. Band 8 is in low gain for all routine acquisitions as the noise level in this band is such that high gain provides very little improvement in performance. were used to evaluate the gain settings. 20002 gain change locations starting moving as necessary to avoid impacting high priority or potentially clear scenes. et. The at-   56  . 1999) "The 8-year monthly averages of AVHRR visible and near infrared planetary reflectance measurements from the AVHRR data set. 2001 gain changes for scenes in Northern Africa are now made several rows earlier so they occur over the Mediterranean. Bands 8 set to low gain 5. When over the U. set to high gain b.7 set to high gain b.S. Water/Coral Reefs: a. which varies by ±1. Bands 5. 2000. This implementation is currently under review and the band 8 gain settings may be altered in the future. Bands 1-5. Several additional changes have been made to the gain strategy plan implemented on July 13.3 At-launch Gain Setting Strategy Rationale The following paragraph was extracted from a paper that describes the rationale behind the at-launch gain setting strategy (Goward. On May 8. Visible (bands 1-3). 6. When not over the U.S. Volcano/Night .nighttime imaging (sun elevation < 0) is only routinely performed for sights identified as "Volcano" a. gain changes are made during the cloudiest scene prior to the scene requiring the change. Band 8 set to low gain The actual saturation reflectance corresponding to a given sun angle are influenced by the Earth-Sun distance.

6.8 seconds 3. For each gain state.S. The rational for setting the offset to -4 seconds was: 1.0 km. 1999 (1999/128:22:07 zulu). in which the along track error could reach 2 seconds). For each Landsat WRS scene. so it would occur during the trailing end of the preceding scene. 1999. This change was made permanent starting on August 2.4 Gain Change Location Rationale The gain change commands at the time of launch were issued at the start time of a scene.   57  . This was initially incorporated between July 14 to July 26. an entropy statistic was calculated to determine the potential scene contrast in each setting. The -4 second gain change offset exists for all scenes acquired after August 2. Gain change command execution time: 0. pre August 2. equivalent to 0. 1999) was compiled for the record. Where low gain was found to provide substantially greater scene contrast (e.09 seconds due to orbital eccentricity.1 second/band X 8 bands = 0. the observed spectral radiance was subjected to the two gain states. glaciers in summer).g. which due to various error accumulations was placing the gain change 1 to 2 seconds into the start of the scene. 2. Scene center-to-center time variance: typically 23.e.92 +/. The MOC was informed that the U.0.satellite planetary reflectance was converted to at-satellite radiance. based on the solar zenith angle at the time of satellite overpass.4. Payload command timing error: up to 1-second quantization error because the SCP executes commands on integer second boundaries. 4. A list of scenes with leading edge gain changes (i. For all other cases the high gain was selected. this setting was selected.2 seconds (except on days following delta-V orbit correction maneuver. but it was the best compromise to meet the requirement to minimize gain changes while providing generally high quality measurements". This decision process will no doubt yield less than optimum results for some applications. ground processing system could not handle this band change location. A scheduler modification was subsequently made to move the gain change commanding back by 4 seconds. 1999. Orbital along-track position uncertainty: typically < 1.

6. one can quickly ascertain Landsat's unique niche. Upon careful examination of this chart. Landsat's retrospective archive stretches back 25 years. Landsat 7 occupies a unique niche between these two extremes. Listed with each sensor is its spatial resolution. near IR. which was recently launched by Space Imaging. and radiometric calibration accuracy. and thermal IR regions). spectral coverage. and accurate radiometric calibration.Landsat Sensing Characteristics Relative to Other Satellite Systems   58  . large footprints. and panchromatic snapshot sensors such as IKONOS. As depicted in Figure 6. On the other end are high resolution.10 . which include repetitive. visible. and global coverage at high spatial resolution in all four passive optical regions of the electromagnetic spectrum (i. small footprint. In addition. No other sensor can match Landsat's uncommon characteristics. spatial resolution. broad-area. Figure 6.    The horizontal bars represent proportionately scaled footprints of the sensors on the left. short-wave IR.e. The right side of the chart lists the sensor's temporal resolution and pointing capability.10.5 Radiometric Characteristics The ground footprint. On one end of the scale are the low-resolution. and the spectral channels of today’s sensors characterize the civilian space-based remote sensing industry. multi-spectral sensors such as NOAA's polar orbiters that have one-kilometer resolution and a 2000-kilometer swath or footprint.

Eliminating their presence from data products is addressed in a later chapter on Level 1 processing. and downlink errors. It is manifested in a noise pattern commonly known as banding. Dropped lines and inoperable detectors also exist as a result of decommutating errors and detector failure. are subsequently removed in a post-processing step using cosmetic algorithms. As a descendant of the Thematic Mapper. include banding and striping. memory effect. The goal of the ground system is to remove image artifacts prior to radiometric processing. 7. if they exist. Remnant artifacts. another may change from high to low).7. It can be observed as alternating lighter and darker   59  . Remnant artifacts. modulation transfer function.3 Memory Effect Thematic Mapper data was rife with the artifact known as ME. A discussion of each of these effects and characterization methodologies follows. the ETM+ generates data with similar characteristics. the exact location of the transition from one SCS state to another within a scan line is determined.2 Scan Correlated Shift Scan Correlated Shift (SCS) is a sudden change in bias that occurs in all detectors simultaneously. Second. First. A proactive approach is in place for Landsat 7. Not all detectors are in phase. This must be done to all scenes that require removal of SCS. All detectors shift between two states that are constantly time varying. Third. some are 180 degrees out of phase (i. The known ETM+ image artifacts are scan correlated shift. and coherent noise. when one detector changes from low to high. these effects were ignored or artificially removed using cosmetic algorithms such as histogram equalization during radiometric processing. SCS levels are obtained for each detector in each line of the scene. dead or dying detectors. or slowly time varying on the order of days to months. In the past. The bias level switches between two states.e. (assuming that a continuous flow of data from the detectors is available). Characterization of SCS is performed at three levels. 7. the value for the SCS levels is calculated for each detector.1 Overview Analyses performed on Landsat 5 data over the years has revealed the existence of imperfections or image artifacts caused by the instrument's electronics. which may exist. Measurement of SCS levels is affected by another instrument artifact known as Memory Effect (ME).

7. Therefore.k(1 . Since the response will be exponential in form. and linearly regressed to determine the model parameters.horizontal stripes that are 16 pixels wide in data that has not been geometrically corrected. the data are manipulated by subtracting out the appropriate bias level. but the more dominant components are not. linearized via the natural logarithm. The most persistent and dominant component is scan-free and has a varying frequency. ME have historically been the cause of significant error in calibration efforts since its effect on IC calibration data is scene dependent. These stripes are most intense near a significant change in brightness in the horizontal (along scan) direction. It is difficult to characterize and correct for as a consequence. ME has been modeled as a simple first order linear system and only three model parameters need to be identified to characterize it (actually one of the three is simply detector bias).' Another artifact known as Scan Line Droop' was originally thought to be a separate phenomenon.   60  . such as a cloud/water boundary.exp (t/tau)) where: g(mf) = the ME pulse response b = detector bias k = ME magnitude tau = the ME time constant Although the exact approach taken to characterize ME is dependent on data type. This directly corresponds to time constants of approximately 1100 minor frames that have been derived from night scenes. It is primarily due to a portion of a feedback circuit that contains a resistor/capacitor combination with a time constant of approximately 10 ms. Because of this. ME is known to be caused by circuitry contained in the pre-amplifiers immediately following the detectors in the instrument electronics. it was formerly termed 'Bright Target Saturation' or 'Bright Target Recovery. with a maximum amplitude of 1 DN. coherent noise (CN) manifested itself in various ways. there is a bursting broadband component that is also not locked to the start of the scan. The power of this component varies strongly even within a scan.4 Coherent Noise In TM reflective band data. and nearly absent in the Cold Focal Plane. response of the detector to some type of pulse is measured and averaged over many scan lines. Some CN components are locked to the start of the scan. but has since been shown to be simply another manifestation of ME. It is present in Bands 1 through 4 of the Primary Focal Plane. In addition. Because of its nature. g(mf) = b . Analysis of a swath of L5 TM night data showed the frequency of this component generally increasing as a function of time with episodic and strong jumps occurring at lamp state transitions.

These parameters will be obtained per scan. the frequency. and the total power in the line in excess of that of the background noise. Finally. Analysis of the swath of night data previously referred to. Dropped lines are characterized by examining the raw input image for the 0-255 fill pattern on a line-by-line basis. scan direction. Interestingly. Thus.1 inverse minor frames (imf). the background noise must be characterized. it is necessary to characterize the continuum of the power density spectrum before identifying CN components.1-. limits. 7.if data are missing from part of a minor frame.6 DN. CN components and the background noise (DN) correlates positively with SCS state. During LPS processing a fill pattern is used to distinguish good data from bad data.3 DN over a timescale of about one thousand scan cycles.2 DN. which is referenced during level 1 processing. It also showed a dependence on lamp state.In addition. Its amplitude may reach up to . It is quite strong in one detector. which provides no change in output DN for changes in input radiance it is also important to determine when a detector channel has fallen out of acceptable performance. this component is difficult to filter out. Consequently. 7. Another scan-locked component appears in only a few detectors. Known inoperable detectors are excluded from this operation. reveals its power to decrease exponentially from an amplitude of 0. Odd detectors are filled with zeros.6 Inoperable Detectors Image artifacts caused by dead or dying detectors require characterization. the power of this component varies significantly. Dropped lines can thus be entire or partial scans.6 DN to an amplitude of 0. having an amplitude of 0. the entire minor frame is filled. During processing filled minor frames are flagged in the label mask. it is assumed that the continuum depends only on detector. and SCS state. Analysis of the night data swath showed a strong dependence of the continuum on detector and SCS state. A starting point for such a determination is a test to see if   61  . In addition to the obvious case of a "dead detector. Filled minor frames are tallied by scan and for the entire image and compared to LPS counts reported in their metadata. but the effect is weak. while even detectors are filled with 255s. its peak frequency varies widely over a range of 0. A common scan-locked component manifests as a spike every 16 minor frames. Data filling is performed on a minor frame basis ." one.5 Dropped Lines Dropped lines occur in 0R data due to decommutating errors in the raw data stream ingested by Landsat Processing System. if not all. the power of most. The set of parameters characterizing a CN component in a detector will be the phase relative to a reference minor frame. frequency) are stored in the metadata that accompanies all subintervals transferred to the LP-DAAC. In order to obtain the total power in excess of the background noise. Statistics on dropped lines (occurrence count. Thus.

7. in homogeneous or "smooth" image regions. the other operates only on the output of FASC scenes. This specification provides a detector to be classified as degraded. One piece takes the indications of dead detectors from the output of performance in the IC to determine detector "aliveness".g. The operation is cosmetic in nature as it recognizes and removes striping patterns left over from an ineffective ME correction process.each detector meets the performance criteria established in the Landsat-7 System Specification. Thus if the region past the bright target is uniform. This latter piece has two portions: one of which can operate on any scene using the shutter data.3. If any portion of the algorithm detects a change in performance (e. a new dead detector.) The filtering operation produces an output image where banding has been removed. if its Signal to Noise Ratio (SNR) or dynamic range are below the specification levels.1. and it's amplitude.1. typically on the order of 1 to 2 counts.   62  . is a sometimes visible noise pattern caused by memory effect.) The result is that scans in one direction will be noticeably darker than adjacent scans in the opposite direction. the filter is adaptive so that it only operates on those portions of the image where banding is detectable (i.8. data values obtained from the sensor will be slightly lower than corresponding values obtained on the following scan (since the following scan is in the opposite direction and therefore. After scanning past a bright target such as clouds or snow.e. This is a distributed algorithm that evaluates the outputs of several of the characterizations. noise level below spec or dynamic range below specs) flags will be set for an analyst to examine the results and potentially update the parameter file. has yet to encounter the bright target.7 Banding Banding or "scan-to-scan striping". 3. An overall figure of merit is also calculated. to assess the dynamic range of the channels. detector response is reduced due to memory effect. A second piece takes the outputs of random noise characterization to determine "in-spec" or out of spec behavior of the detector noise levels. Banding characterizations are performed after artifact removal and radiometric correction. as well as a difference image that gives an indication of where banding was detected. The final piece of the algorithm takes the output of the current gain selection and the currently indicated saturation bins. The banding pattern is very small in intensity. Para.7. Because of the small amplitude inherent to banding. The approach employs an algorithm that applies a filter optimized for detection of the banding pattern.

The spectral filters. this 5% requirement is supported by a monitoring and calibrations program. effectively de-striping the data. 8. i. This is achieved by linearly adjusting the 1R data to match the means and standard deviations of each detector to a reference detector. The cold focal plane assembly has a filter housing   63  . primary.1 Spectral Characterization The measured wavelength locations of the ETM+ spectral bands are compared to Landat 5's TM in Table 8. a post processing characterization and removal process is in place should a cosmetic fix become necessary.1 The spectral bandwidths are determined by the combined response of all optical path mirrors (i.1 Radiometric Calibration Overview A major objective of the Landsat-7 program is to upgrade the radiometric quality of the data to be commensurate with the other sensors in the Earth Observing System (EOS). located immediately in front of each detector array. and these are included in the process of generating the applied gains and biases. Pre-launch. and the implementation of any necessary changes to the ground processing of the data.2 Pre-Launch 8. Its source can be traced to individual detectors that are calibrated incorrectly with respect to one another. and instrument characterizations. This actual processing algorithm applies the relative gains and biases calculated by the histogram analysis performed on every image just prior to radiometric correction. The prime focal plane assembly has a filter housing that contains filters for bands 1 through 4 and the panchromatic band.e. Unlike its predecessors. and the individual detectors.2.7. scan line corrector. the spectral filters. the striping characterization and correction should not be required in routine processing. are the dominant items that establish the optical bandpass for each spectral band. 8. secondary. As detector to detector variations are already explicitly taken into account through the generation of relative gains and bias from histograms. a specific goal of the Landsat-7 program is to achieve radiometric calibrations of the data to ± 5% uncertainty over the 5-year life of the mission.e.8 Striping Striping is a line-to-line artifact phenomenon that appears in individual bands of radiometrically corrected data. scanning). the mission design supports this requirement through hardware design changes. Nonetheless. is intended to remove the detector to detector variations in gain and offset. Post-launch or on-orbit. or to the average of the detectors. The application of the calibration coefficients to the ETM+ data. the generation of the level 1R data.

0.35 0. 2 100-watt lamps.that contains filters for bands 5 through 7.63 . the new filters have shown significant improvement in band edge responses as compared to Landsats 4 and 5.4 .1.78 .10.2.4 . In addition. and 4 25-watt lamps. The first .90 1.53 . 6 45-watt lamps.45 0.52 0.63 .09 0. It provides radiance levels covering the full dynamic range of the instrument in all bands.75 12.0. The SIS48 was used for monitoring the radiometric calibration of the ETM+ five times during instrument and spacecraft level testing. and 10 8-watt lamps.90 1.   64  .76 .5 2.52 . 8.69 0.2.55 .60 0.1.75 12.69 0.61 0.55 . The ETM+ filters were made using a process called ion assisted deposition (IAD) which presumably makes the filters resistant to this phenomenon.52 Band 8 Plot Data N/A . During SIS calibrations the Bench Test Cooler (BTC) was used to maintain the temperature of the Cold Focal Plane at 105°K. Table 8.a 100 cm source (SIS100) is equipped with 18 200-watt lamps. The second source is a 122 cm (48") SIS with 6 200-watt lamps.0. The SIS100 was used to perform the primary radiometric calibration of the ETM+ in August 1997 and was also used for the pre-launch calibration of AM-1's Moderate Resolution Imaging Spectroradiometer (MODIS).5 2. and at least 10 usable radiance levels for each band for each gain state.Half Maximum Band Band Band Band Band Band 2 3 4 5 6 7 Plot Plot Plot Plot Plot Plot Data Data Data Data Data Data 0.90 A discrete spectral shift occurred on Landsat 5 TM has been largely attributed to filter outgassing.10.2.08 0.0.35 Sens or TM ETM + Band 1 Plot Data 0.1 TM and ETM+ Spectral Bandwidths Bandwidth (µ) Full Width .52 .45 0. This was the only one of the three temperature set points for the cold focal plane that could be obtained in ambient pressure and temperature conditions..2 Radiometric Calibration Reflective Band Calibration and Monitoring Two spherical integrating sources (SIS) were used to calibrate the ETM+ prior to launch.

and the intercepts are the biases.b). The Landsat Project Science Office (LPSO) will review the various integrating sphere calibrations and their effective transfer to the ETM+ before deciding which calibration should go the IAS to represent the pre-launch IAS. (G(d.b). Thermal Band Calibration The radiometric calibration of band 6. the thermal band. is fundamentally different than the reflective bands as the instrument itself contributes a large part of the signal.The calibration data reduction is performed as follows: The slopes of these regression lines are the responsivities or gains. A model of this temperature dependent instrument contribution has been developed by SBRS. The calibration for band 6 is formulated as:                             65  . B(d.

a collimator with selectable sources at its focus.e. Radiance is defined as the energy flux per unit projected area per unit solid angle leaving a source.3 ETM+ Spherical Integrating Source   A spherical integrating source is a hollow sphere with the entire inner surface uniformly coated with a material.SIS in clean room distributed in all directions equally. A perfect diffuse reflector can behave like a perfect (i. The results of this calibration are nominal gains and biases for band 6. The gains and biases are included in the CPF as pre-launch values for band 6. The radiometric calibration of the thermal band occurs during thermal vacuum testing.2. which has a high diffuse reflectance. blackbody sources will be used in the TMC. Each SIS is calibrated by SBRS to National Institute of Standards and Technology (NIST) traceable standards of spectral irradiance. A Lambertian source is a source whose radiance is independent of viewing angle. and the emissivity adjusted view factors (a(j)) for the various internal components of the ETM+ that affect the band 6 calibration.The pre-launch calibration of band 6 is primarily a calibration of this model. The basic concept behind the spherical shape is that light from the internal source has a chance to perform multiple bounces thereby randomizing its original direction before it exits a small aperture. During the band 6 calibration. In addition. During this test the ETM+ is aligned to the Thematic Mapper Calibrator (TMC).1 . Lambertian) diffuse source which means energy is Figure 8. or a surface. 8. The band 6 detectors' responses to combinations of various TMC blackbody and instrument temperatures are used to calibrate the instrument and to refine emitted radiance contributions from various internal ETM+ components. EOS cross calibration activities   66  . The sphere's interior coating is designed to have a very high degree of diffuse reflectance.

8. University of Arizona. via independent analysis. the Partial Aperture Solar Calibrator (PASC).3. while the PASC and FASC are mainly useful for the reflective bands. The light is directed out of the shutter flag and onto the focal planes by additional optics in the head of the instrument.3 Post-Launch The post-launch radiometric calibration of the ETM+ is accomplished by regularly examining the instrument's response when illuminated by known sources that are relatively stable. the accuracy of the calibration using on-board sources. namely.1 Internal Calibrator The IC consists of a shutter flag.include comparison of the SBRS radiometric scale to the NIST. from the focal planes. The IC is useful for calibrating all ETM+ bands. The light separated for each band is aligned so that it impinges on the appropriate detectors. The Landsat Transfer Radiometer (LXR). is used to monitor the output of each sphere during each calibration and calibration check. located immediately in front of the prime focal plane. into a sapphire rod contained within the flag. The two IC lamps are situated near the base of the internal calibrator flag. Ground look calibrations are occasionally performed to confirm. Changes to the ETM+ calibration have occurred since launch and can be viewed using this graphical timeline (Table 8. At the end of each scan the shutter blocks light from the Earth.2). In addition. Light from either or both lamps is directed through optics at the pivot point of the flag. a visible and near infrared radiometer designed for stability by NIST.   67  . to the detectors. The shutter flag. The ETM+ has 3 on-board calibration devices. This radiometer has also been calibrated by NIST to provide an independent check on the radiometric calibration of the two sources. NASA's Goddard Space Flight Center. and Japan's National Research Laboratory of Metrology (NRLM) radiometric scales. oscillates in synchronization with the scan mirror. the shutter flag relays light from the IC lamps and blackbody. 2 tungsten lamps. the Internal Calibrator (IC). and a blackbody source.  8. This rod transfers the light up the shutter flag and splits it into separate paths for each of the spectral bands. and the Full Aperture Solar Calibrator (FASC).

whereas TM used radiance stabilization in the primary mode. Low gain. radiation from the blackbody is reflected off of a toroidal mirror on the flag. while the TM's used fiber optics in an attempt to improve the uniformity of the calibration flux at the focal plane. is coated with a high emissivity paint and acts as the second source for band 6 calibration. such that 4 lamp states are possible (both "off" [0.2 .1]).0] or both "on" [1. 12/7/96. though the lamps are nearly identical in terms of current and voltage ratings. When the shutter flag passes in front of the primary focal plane. both lamps "on" will saturate some bands particularly in high gain mode.   68  . exclusive of the area where the toroidal mirror is located. This portion of the shutter flag is also instrumented with a thermistor. (3) the control circuit for ETM+ uses voltage regulation in the primary operation mode. and voltage regulation in the backup mode. The IC was designed to have one lamp produce a usable signal in all bands. (2) a more compact filament results in a higher flux incident on the IC optics. The portion of the shutter flag imaged by band 6.1] or [1. in 5 principal ways: (1) the ETM+ uses 2 lamps (4 states) instead of 3 lamps (8 states). (4) ETM+ uses sapphire rods to transmit the energy from the base of the flag to the head of the shutter flag.ETM+ Band 4. although similar to the IC on the Landsat-6. (5) the ETM+ does not retain the lamp sequencer used on TM to automatically cycle through the lamp states. into the aft optics of the ETM+ and onto the band 6 detectors. IC data. Figure 8. one "on" [0. The IC blackbody is situated off the optical axis of the instrument. resulting in quasi-constant power being supplied to the lamp. The blackbody has three set point temperatures namely. 37°C and 46°C. The ETM+ IC.0]. Detector 1. 30°C. primary lamp. Each lamp can be commanded "on" or "off". differs from the IC's on Landsat-4 and Landsat-5. Note.The IC lamps are supplied with a regulated voltage across a combination of the lamp and a resistor.

Figure 8. is specified to be 23. the shutter flag oscillates in synchronization with the scan mirror. This paint was selected for it’s near Lambertian properties.When the ETM+ is operating. When stowed the panel rests adjacent to the stow cover which reduces the exposure of the panel to contaminants and UV radiation.2 msec. The center 51 cm of the FASC panel is painted with the classic formulation of YB71. solar irradiance and geometry conditions. this device behaves as an independent. With known surface reflectance. for the 30 meter channels. and apparent stability in a space environment. sunrise on panel) to about 67° zenith angle.2 Full Aperture Solar Calibrator The Full Aperture Solar Calibrator (FASC) is a white painted panel that is deployed in front of the ETM+ aperture and diffusely reflects solar radiation into the full aperture of the instrument as illustrated in Figure 8.ETM+ FASC in deployed position in front of aluminum honeycomb paddle on ETM+ aperture. has a width of approximately 40 pixels (Figure 8. a motorized arm. The size of the shutter flag and its speed of movement combine to provide obscuration of the light to each detector for about 8. When in the calibrate position the angle between the sensor nadir vector. high reflectance. The image will increase in brightness along track as the solar   69  . The device consists of an octagon shaped. the light pulse for the reflective bands. full aperture calibrator.3 . The relative azimuth between the nadir view vector and the solar illumination varies across the same range.5°.3. Below 67°. an inorganic flat white paint designed for spacecraft thermal control. and the panel normal. For band 6. the instrument begins to shade the panel. to an inclined position in front of the ETM+. Depending on the time of year the solar azimuth angle with respect to the velocity vector of the ETM+. the panel can be illuminated by the sun from 90° zenith angle (i. the motor rotates the panel from its stowed position away from the ETM+ aperture. 8.2). or 750 pixels. In use.3. the calibration signal is similar with the blackbody pulse about 20 pixels wide. ETM+ image data acquired with the FASC will appear to be an essentially flat field with vignetted cross track edges. varies from 23° to 37°. On command.e.

the resolution being limited by diffraction from the small apertures. Pre-launch BRDF measurements indicate that the radiance change across the scan. Although the spacecraft solar panel undergoes a period of thermal instability during sunrise.4 . which resulted in a nominal declination angle of the PASC (relative to spacecraft nadir) of 20 degrees.zenith angle (SZA) on the panel decreases (roughly at 1/cos(SZA)). a 45° mirror.3 Partial Aperture Solar Calibrator The Partial Aperture Solar Calibrator (PASC) is used for calibrating bands 1-5. should be a 1% effect across the full scan assuming the aperture is filled. The four facets are duplicated to account for angular variations of Sun position with season. and the view azimuth angle varies by +/. Each set (or facet) consists of an uncoated silica reflector. and an aperture plate with a precision drilled small aperture (~4 mm) (Figure 7). This diffraction effect is wavelength dependent. the design nominally fills the aperture for the central ~50% of the scan line. The PASC will generate a reduced resolution image of the sun. If the nadir viewing pixel has the nominal 23. at least one facet will reflect sunlight directly into the ETM+ aperture and the ETM+ will image the sun. sufficiently to bring it into the ETM+ dynamic range. It is attached to the ETM+ sunshade and permanently obscures a small portion (~0.5%) of the aperture.30°.e.Simulated ETM+ PASC scene. an analysis of the resultant spacecraft jitter has shown minimal impact (< 1%) to the acquisition of PASC data. It consists of four essentially identical sets of optical elements each in a slightly different orientation. the view angles to the FASC panel change. have revealed a design misalignment. then at the extreme ends of the scan.3. space vehicle sunrise) in the vicinity of the north pole.1% effect. Specifications require the FASC to fill the ETM+ aperture for the central 1000 pixels (approx 1/6 of each scan line). They are oriented such that in any given orbit. This increase in declination effectively forces the Spacecraft to acquire PASC scenes earlier in its orbit (i. For   70  . as the satellite passes out of solar eclipse (i. this translates into a 0. closer to spacecraft sunrise). 7 and 8 and consists of a small passive device that allows the ETM+ to image the sun while viewing a 'dark earth'. Recent SBRS measurements of the alignment between the PASC and scanner assembly. As the mirror scans. The combination of the small aperture and the uncoated silica reflector reduces the signal amplitude Figure 8. versus the prescribed 18 degrees.e. 8.5° view angle and a 0° view azimuth angle. Across the central 1000 pixels. the view zenith angle increases by about 1°.

in band 1 the blur will extend across about 7 pixels (at the first dark ring of the diffraction pattern). 8. and 410 pixels for band 8. These combined effects of oversampling. it is anticipated that most uniform portions of the solar disk center will be approximately 200 pixels in width for bands 1-5.e. skewed image of the sun as seen in Figure 8. In addition to the blur.date image was acquired by the satellite Processing Date .5 degrees/minute.039 degrees (16 * .2 are the actions taken by the Landsat Project Science Office (NASA/GSFC) and the Image Assessment Team (USGS/EDC) to provide the best possible calibration for Landsat ETM+ imagery. the ground is normally scanned at 16 instantaneous field of views (IFOV's). this date should be close to the date the scene was ordered and received by the customer   71  . By comparison. every day. Within a PASC processed image.0024 degrees) per 72 msec scan or ~32.3. PASC calibration is performed once a day.example. Definition of Terms: Acquisition Date . on the orbit specified by the IAS. and a nonorthogonal scan pattern. which is not perpendicular to the motion of the sun .4. is the scanning direction. The IAS orders the resulting data from the EDC-DAAC for calibration processing and assessment. the image will be elongated in the "along-track" direction. This equates to . The "along-track" movement across the solar disk can best be expressed in terms of the spacecraft pitch rate (i. 7.6 degrees/minute). Thus the sun image will be oversampled along track by a factor of about 9. One other contributor to the rendition of the solar image. and in band 7 the extent is about 32 pixels.date image was processed to Level 1 by the Processing System.the angle between the two can be as small at 45 degrees. produce an elongated.4 ETM+ Calibration Actions Presented in Table 8. 360 degrees in ~100 minutes or ~3. 105 pixels for band 6.

31 radiance units B6 out of calibration calibrati on correct ed Lmin/Lmax update all all use first set of Lmin/Lmax's to scale Level 1 DN to radiance use second set of Lmin/Lmax to scale Level 1 DN to radiance Summary of Required Actions Problem/Update Details/For more details. resulting in an intermediate B6 Gain Swap gain state.usgs. For B6 Calibration users with data processed before 21 DEC 2000. Independent calibration teams discovered an offset error in the B6 calibration.. subtract 0.. the scaling factors Lmin/Lmax update were adjusted beginning 01 Jul 2000 to make better use of the dynamic range. subtract 0.gov/news/ and USGS: https://landsat.31 radiance units for calibrated imagery calibration corrected B6 B6 Gain Swap Period 6 11/24/99 to 12/08/99 B6 out of calibratio n bias error in calibration procedure. Using updated information on the pre-launch calibration.Table 8.31 W/m2 sr um from the radiance image for calibrated data.php   72  .gsfc. An error caused the two B6 gains to be swapped.2 Calibration Action Time Line Problem/ Update Bands Affected Acquisition Dates 1999 J u l B6 Calibration 6 all A u g S e p O c t No v De c Processing Dates 2000 A Jan Feb Mar p r M a y J u n J u l A u g S e p O c t N o v D e c J a n F e b M a r A p r 2001 M a y J u n Jul bias error in calibration procedure. This was corrected through processing on 21 DEC 2000.gov/about_Landsat_Updates. Links to Landsat Updates at NASA: http://landsat. Attempts to correct the problem were implemented then lost and implemented again on 01 Jul 2001.nasa. subtract 0.

et.4.8. there is a limited overlap period when more than one of the sensors is operating at the same time. ETM+).e. the Landsat-7 ground track drifted slowly relative to the nearly fixed Landsat-5 pattern. This unusual and valuable opportunity was specifically designed to facilitate the establishment of data consistency between the Landsat ETM+ and TM sensors. including the development of a stable sensor (i. 1999 placed the spacecraft temporarily in an orbit very close to that of the Landsat-5 spacecraft. Consistency between the Landsat sensors starts with sound calibration of the individual sensors. In order to benefit from this data record. Typically. Additional processing steps to retrieve surface parameters such as reflectance and temperature then become possible. During the tandem configuration period when there was useful   73  . Such an overlap period with Landsat 5 was designed into the initial phases of the Landsat-7 mission. At this altitude. 1999 when the tracks were almost exactly the same. al. but with a temporal offset on the order of 10 to 30 minutes. The mean altitude of Landsat-7 was 699 km. The Tandem Configuration The launch of Landsat-7 on April 15. and on-orbit calibration. In anticipation of a successful Landsat 7 mission. 2001). ETM+ cross-calibration with earlier Landsat sensors begins by making use of near-simultaneous imaging of common Earth surface targets. renewed efforts were made to ensure radiometric calibration across the Landsat series of sensors and with other Earth observation sensors. steps are needed to ensure that the data are self-consistent and not significantly affected by artifacts of the various Landsat sensors. This material was extracted from a paper that covers the subject in greater detail (Teillet. yielding image data at the top of the atmosphere in physical units. 6 km below the 705-km mean altitude of Landsat-5. A critical step in such a process is sensor radiometric calibration to an absolute scale.. detailed prelaunch characterization.4 TM Cross Calibration 8. Post-launch radiometric calibrations are based on reference to onboard standards and ground-based test sites. The resulting opportunity for radiometric cross-calibration between ETM+ and Landsat 5's TM is the main subject of this chapter. The key period for the tandem configuration was June 1-4.1 Introduction Calibrating the Landsat Data Record The Landsat data record is important for terrestrial remote sensing and global change research because it covers a 30-year time when significant anthropogenic terrestrial change has occurred.

  74  . Subsequently. in cooperation with Space Imaging EOSAT and international ground stations.3). the Landsat-7 orbit was changed for nominal operations such that its 16-day repeat coverage cycle is now offset from that of Landsat-5 by 8 days. the number of useful tandem data scene pairs is roughly estimated to be on the order of 400 scenes. Given cloud cover and possible problems with data reception and recording. image sequences corresponding to 791 matching scenes were recorded by both the Landsat-7 ETM+ and.overlap in coverage between the two sensors. the Landsat-5 TM (Table 8.

Libreville. Canada RSA Saudia Arabia (for SI/Dubai) NOK SI/EOSAT. Norman. Prince Albert. Gabon FUI ESA. Cuiaba. Brazil ASA ACRES. Australia .Table 8. Argentina CUB INPE. Oklahoma JSA Johannesburg. South Africa KIS ESA. Sweden LBG DLR. Fucino. 1999) Stations June 3 3 3 2 2 3 3 2 2 3 3 1 2 2 3 2-3 3 2 3-4 2   P 6 6 6 15 15 22 22 31 31 38 38 40 47 47 54 95 102 104 111 152 R 2129 5771 6771 1244 2745 1043 2649 740 2546 639 2540 2538 430 2530 425 6587 6983 6282 6484 Station GNC CUB COA GNC NOK GNC NOK PAC NOK PAC NOK NOK PAC NOK PAC ASA ASA ASA ASA RSA June 3 3 2 2 2 3 3 3 3 2 2 3 3 2 3 2 3 3 2 2 Path 159 159 168 168 168 175 175 175 175 184 184 191 191 200 207 216 223 223 232 232 Row 69-78 19-27 62-83 19-26 23-42 62-85 22-44 44-77 14-24 17-43 17-40 19-24 63-76 60-86 68-98 54-85 66-97 Station RSA JSA KIS RSA JSA KIS FUI RSA JSA FUI LBG KIS FUI FUI FUI CUB CUB COA CUB COA 75  PAC CCRS.3 Landsat 7 ETM+ and Landsat 5 Tandem Data Coverage Tandem Scene Coverage (June 1-4. Gatineau. Alice Springs. Canada COA Cordoba. Italy GNC CCRS. Kiruna.

4 Characteristics of the Two Tandem Data Sets Railroad Valley Playa Niobrara. 8.4 km 106 km by 66 km * A0D550 represents aerosol optical depth at 0. a team from South Dakota State University acquired the same types of ground reference data at a grassland test site in the area of Niobrara.23 ° 28. The RVPN test site is a dry-lake playa that is very homogeneous and consists of compacted clay-rich lacustrine deposits forming a relatively smooth surface compared to most land covers.4. Similarly.The cross-calibration methodology documented here is applicable to tandem image pairs and presents specific results for two different pairs of nearly coincident matching scenes from the tandem configuration period. Nebraska Image Date June 1. when a team from the University of Arizona made measurements of surface spectral reflectance and atmospheric aerosol optical depth the same day.4 shows both Landsat image pairs.760 km ETM+ A0D550* 0.7 km by 4.425 km 0. The formulation includes adjustments for differences in illumination regimes as well as for differences in spectral response profiles between the two sensors.56 km East 18.550 micrometers Table 8.2 provides information on the characteristics of the two data sets and Figure 8.60 ° TM Solar Zenith Angle 27.059 TM A0D550* 0. Nebraska (NIOB) on 2 June 1999.4. 1999 June 2.059 Area Common to ETM+ & TM 10. Nevada on 1 June 1999. 1999 WRS Path/Row 40/33 31/30 Landsat 7 Offset from WRS 76.67 ° Terrain Elevation 1. the day of the tandem Landsat overpasses for that site.15 km East ETM+ Data Level Level-0R Level-0R TM Data Level Level-0 Level-0 ETM+ Solar Zenith Angle 24.1035 0.2 Tandem Data Sets Selected for Analysis Attention was focused on two particular tandem image pairs for cross-calibration methodology development and analysis because of the availability of ground reference data. Table 8.28 ° 26. The main results consist of TM responsivities in the six solar reflective spectral bands referenced against well-calibrated ETM+ responsivities in corresponding spectral bands. The NIOB test site is characterized primarily by grasslands grazed by cattle and by a smaller proportion of agricultural crops. Both Landsat sensors imaged the Railroad Valley Playa (RVPN).   76  .1046 0.

1999).2. the Landsat-7 and Landsat-5 sensors differ in their along-track and acrosstrack pixel sampling. there is variation in the ETM+ scanning pattern and its effect on the scan line corrector due to the lower-than  77  . radiometric. The solar illumination geometries are very similar (within three degrees). and relative azimuth angles between solar and satellite directions do not differ significantly from one Landsat overpass to the other. Nevada WRS For Landsat sensor image data pairs 40/33 acquired June 1. slight mismatches will arise in the imagery because of the altitude difference.8. and spectral considerations to be addressed. Because data acquisitions were only 10 to 30 minutes apart during the tandem configuration period. 1999 configuration period.3 Cross Calibration Methodology The cross-calibration methodology assumes that the Landsat-5 TM calibration is to be updated with respect to the Landsat-7 ETM+ sensor.Landsat 5 and 7 Cross Calibration illumination and observation angles. acquired during the tandem Nebraska. 1999 and of Niobrara. there are also across-track differences in the ground coverage.4. In particular. it is assumed that the surface and atmospheric conditions did not change significantly between the two image acquisitions. Geometric Matching Geometrically. Cross-calibration methodologies in general should consider adjustments as appropriate for bi-directional reflectance factor (BRF) effects due to differences in Figure 8.5 .. For the same reason and because the ETM+ scan time is slightly longer than the specification. the expectation using bands 5. Nevertheless. satellite zenith angles are predominantly near-nadir. is that such BRF adjustments are not necessary. WRS 31/30 acquired June 2. In addition. along-track gaps between scans are longer than they are for Landsat-7 ETM+. Due to wearing of the bumpers used by the Landsat-5 TM scanning mirror. there are geometric. Data Sets of Railroad Valley Playa. which serves as a wellcalibrated reference sensor with a radiometric calibration uncertainty of ± 3 percent (Barker et al.4.

TOA radiances L*i (in Watts/(m2 sr µm)) are related to image data by (3) where:   78  . The zeroradiance biases are based on dark current restore values computed on a line-by-line basis. Normalized and bias-corrected image values are then given by (2) Thus. the analysis approach was developed to make use of image statistics based on large areas in common between the image pairs. In spectral band i. for each particular scene in the case of TM and for many scenes in the case of ETM+. The normalizations are with respect to the band average and the process is not expected to bias the cross-calibration. Therefore. Quantized levels of Q = 0 and Q = 255 are excluded to avoid saturation effects.nominal orbit during the tandem configuration time period. These considerations make it very difficult to establish sufficient geometric control to facilitate radiometric comparisons on a point-by-point and/or detector-by-detector basis. Radiometric Formulation Radiometric raw data are assumed (Level 0 for TM and Level 0R for ETM+). the image quantized level Qi (in counts) is related to top-of-atmosphere (TOA) radiance Li* (in Watts/(m2 sr mm)) by (1) Gi is band-averaged sensor responsivity (in counts per unit radiance) is the zero-radiance bias (in counts) in spectral band i. Radiometric detector normalizations based on full-scene statistics are applied in each spectral band.

and (4) yields (5) There are two advantages to using reflectance instead of radiance. The other advantage is to compensate for different values of exoatmospheric solar irradiance arising from spectral band differences.5 percent effect in the ratio of the cosines of the respective angles. (3). Cross Calibration Equation (5) can be defined separately for image data from the Landsat-5 TM ("5") and for image data from the Landsat-7 ETM+ ("7"): (6) (7)   79  . If differences in atmospheric conditions are not a factor.TOA reflectance is related to TOA radiance by (4) where: E0i is the exo-atmospheric solar irradiance in spectral band i (in Watts/(m2 mm)) based on the Modtran-3 spectrum is the solar zenith angle ds is the Earth-Sun distance in Astronomical Units A combination of equations (2). For example.to 30-minute time difference between data acquisitions. then the TOA reflectance comparisons have the potential to yield the best possible calibration comparisons between the TM and ETM+ based on the tandem data sets. One advantage is to remove the cosine effect of different solar zenith angles due to the 10. the three-degree difference in solar zenith angles for the RVPN image pair leads to a 2.

In particular.4. Landsat-5 TM responsivity Gi5 is then given in spectral band i (in counts per unit radiance (CPUR)) by (11) With this updated value of TM responsivity.The combination of equations (6) and (7) yields (8) where the adjustment factor Ai adjusts Landsat-5 TM radiance data for illumination and spectral band difference effects.5. Mi is the slope of the linear equation that characterizes ΔQi5A as a function of ΔQi7 and (9) Where (10) Bi is essentially a spectral band adjustment factor. users can obtain TOA radiance Li* (in Watts/(m2 sr µm)) from raw image quantized levels Qi (in counts) using (12) where ai = 1 / Gi and bi = . Thus. 8.4 Image Processing and Analysis Standard image processing and statistical analysis steps were used to obtain the Mi slopes in equation (8) for use in equation (11). image pairs from the tandem configuration period make it possible to use well-calibrated Landsat-7 ETM+ image data to update the radiometric calibration of the Landsat-5 TM. sub-pixel geometric registration is not critical in this case. given that the ρ*i in equations (6) and (7) are not necessarily the same because of the differences in relative spectral response profiles between corresponding ETM+ and TM spectral bands.Q0i / Gi . As noted before. common to the ETM+ and TM data pairs. but care was   80  . depicted in Figure 8. The ΔQi5A and ΔQi7 for use in equation (8) were obtained from large areas.

Plot of grid cell ΔQ and ΔQ radiometric formulation described earlier. no image cells were excluded on the basis of the jitter exercise. Set up a 5 by 5 grid of contiguous image windows or cells and extract ΔQi means and standard deviations from each of the 25 grid cells for an area common to both the ETM+ and TM image data. (Equation (8)). After preprocessing in accordance with the Figure 8. For the majority of grid cells. yielding 25 subsets of means and standard deviations per grid cell. Repeat step 1 for a series of one-pixel shifts in a 5 by 5 pattern. The value retained is the one obtained for the geometric centre of the jitter pattern. Keep grid-cell mean results only if sensitivity to shifts is low (within one percent). The jitter exercise revealed low sensitivity to possible misregistration between the subscenes selected as areas common to both images in each tandem pair.3 percent in a few cases for the Niobrara scene and just over one percent for one grid cell for the Railroad Valley playa scene. Plot grid-cell ΔQi5A and ΔQi7 means and obtain the slopes Mi (equation (8)). The average coefficient of variation for the RVPN case is 0. Compute ΔQi5A from ΔQi5 using equations (8) .24 percent. the coefficient of variation for the 25 jitter values of ΔQi is a small fraction of a percent. The image processing steps in each solar-reflective spectral band i was as follows. Therefore. Because the quantized levels i5A i7   81  .7).taken to capture the common area as accurately as possible. This "jitter" pattern makes it possible to assess the sensitivity of grid-cell data to these shifts as an indicator of misregistration effects. The lines are linear fits with were plotted to obtain the slopes Mi zero-intercepts.4 shows a plot for the Railroad Valley playa and Table 8.6 .(10) to adjust for spectral band differences and illumination regime differences between acquisitions. Use equation (11) to compute Landsat-5 TM responsivity Gi5 .5 lists the slope results for the RVPN and NIOB subscene pairs analyzed separately and in combination (Figure 8. means combined for the Railroad Valley sub-scene grid-cell means for ΔQi5A and ΔQi5 sub-scene. The additional sets of values resulting from this jitter exercise are not used for any other purpose. reaching 0. Figure 8.

5 Slopes (Mi) and Sensor Responsivities (Gi5 and Gi7) Gi7 Spectral Mi 100 (1-R2) Gi5 (CPUR Band (zero-intercept) (zero-intercept) (CPUR) ) Railroad Valley Playa Sub-Scene Data 1 1. Table 8.6740 3 0.052 7.5525 0.589 7.41   82  .261 2 0.5 also indicates that the Mi slopes obtained for the two different image pairs generally differ by a few percent only.082 5 1.538 0.5879 0.49 1. with the exception of band 4 for the Niobrara case.030 1.589 7.243 2 0.5812 0.589 7.5 1.6662 1.7235 1.9042 4 0.7233 0.22 7. Nevertheless.2 1.80 14.40 1.039 1.2 21.040 0.944 7 0.7 1. the linear fits were forced to have zero intercepts.538 0. which provides some degree of confidence in the cross-calibration methodology.045 0. where R is the correlation coefficient.191 0.191 0.6608 0.26 21.7226 36 1. with the linear fits are low (Table 8. the unaccounted for variances in percent.80 14.496 1. 100 (1R2).00 Combined Tandem Pair Sub-Scene Data 1 1.496 1.5884 2.017 1.8939 4 0.80 14.3 7. No explanation has been found for the greater scatter in spectral band 4.047 1.5).081 5 1.6424 0.6561 3 0.191 0.5509 0.225 1.9050 4 0.6580 3 0.026 1.52 Niobrara.082 5 1.45 1.are bias-subtracted.5659 0.77 1.061 21.538 0.225 1.496 1.014 0.931 7 0.017 0.891 7 0.246 2 0.225 1.60 1. Table 8. Nebraska Sub-Scene Data 1 1.

Deriving spectral adjustment factors for other tandem data pairs would be less straightforward because of the lack of ground reference data. and the surface reflectance. respectively. Standard values were assumed for other atmospheric input parameters (mid-latitude summer profile and continental aerosol model). except in the two short-wave infrared bands where it is larger. Because surface spectral reflectance (Figure 8.8. There are significant differences in relative spectral response profiles between corresponding Landsat-7 ETM+ and Landsat-4/5 TM spectral bands.5 Spectral Band Differences                                  Figure 8. ranging from 3 to 7 percent depending on the band and the test site. these data were used as inputs to a radiative transfer code to compute the TOA reflectance in corresponding solar reflective ETM+ and TM band.4.6 show that the spectral band difference effect is on the order of two percent.Surface Reflectance Spectra.   Figure 8.7 . The effects these spectral band differences have on measured TOA reflectances depend on spectral variations in the exoatmospheric solar illumination. For each test site. the atmospheric transmittance. The lines are linear fits! with zero-intercepts.6). Bi factors were generated for the image data pairs under consideration in this study (Table 8. 1999. but solar illumination angles pertinent to each Landsat image acquisition under consideration were used.Plot of grid cell ΔQi5A and ΔQi7 means combined for the two sub-scenes.6) and atmospheric aerosol optical depth data are available for the RVPN and NIOB test sites on June 1 and 2. which is attributable to the significantly different reflectance spectra of the playa and grassland surfaces. The direction of the effect is also opposite between the two sites in bands 1 and 2. The results in Table 8.   83  .8 .

964 Bi for RVPN 1.981 0.6 Spectral Band Quantities used for Cross-calibration.039 0.118 0. The average difference is 1. although no corroborative analyses have been carried out.969 Ai for NIOB 1.044 1.051 1. NIOB = Niobrara. although based on only 12 spectral band cases. RVPN = Railroad Valley Playa. The jitter analysis indicated a misregistration effect on the order of 0.035 1.932 Bi for NIOB Eo is the exo-atmospheric solar irradiance.2 80.7 also shows differences in TM responsivities if spectral band difference adjustments are excluded (Bi = 1).24 percent and.59 ETM+ Eo(W / (m µm)) 1.997 1. the implication is that cross-calibration comparisons that do not benefit from the surface reflectance spectra and atmospheric optical parameters needed to compute the spectral band difference effect will potentially have an inherent additional uncertainty of several percent.003 1. experience suggests that the other uncertainties are also well within one percent. but greater than that in the short-wave infrared bands. Figure 8. Nebraska. 1999).013 1. Nevada.4 81.4. the overall root-sum-squared uncertainty for the cross-calibration method is approximately ± 3.022 0. Overall. which. Spectral Band 1 2 3 4 5 7 2 1954 1826 1558 1047 217.6 percent.29 TM Eo(W / (m µm)) 2 1968 1839 1555 1054 228. Additional sources of uncertainty include residual geometric misregistration. adjustments for spectral band difference appear to be on the order of two percent or less in the visible and near infrared bands.977 1. If these additional sources of uncertainty amount to a 1-2 percent effect. As noted in an earlier part of the paper. and residual uncertainty from spectral band difference adjustments. The near-simultaneity of image acquisition and the similarity of   84  . The consistency between results from the two image pairs varies from negligible differences in spectral band 4 to almost four percent difference in band 7.053 0.981 0. small changes in atmospheric conditions between tandem image pair acquisitions.7 compares results from the two image pairs with Railroad Valley playa arbitrarily chosen as the reference case.5 percent.017 1.Table 8. A starting point for estimating the uncertainty of the tandem-based cross-calibration method is the ± 3 percent uncertainty of the ETM+ radiometric calibration (Barker et al.026 0.013 1.106 0. and Bi is the spectral adjustment factor. artifacts in the TM image radiometry.020 1. is a measure of the repeatability of the cross-calibration approach. Figure 8.994 Ai for RVPN 0. 8.6 Cross Calibration Results The Mi slopes derived from the two image pairs (RVPN and NIOB) separately and combined were used in equation (11) to generate TM responsivity coefficients.014 1. Ai is the illumination adjustment factor..994 1.

211 0.91 Differe nce Relativ e to CrossCal Gi5 (CPUR) -3..00 1999 SDSU Vicario us Calibra tion Gi5 (CPUR) 1.00% 0.88% -1.6% -4. For spectral bands 5 and 7.6620 0.7 Tandem based Cross-calibration Results and Vicarious Calibration Comparisons 1999 RVPN ETM+ Cross Calibrati on Gi5 (CPUR) 1.221 0.17 Differe nce Relativ e to CrossCal Gi5 (CPUR) -2.9040 0.77 -20% -17% -11% 0.243 0.6561 0.imaging geometry afforded by the tandem configuration are definite advantages in this context.1% 2.7% The TM vicarious calibration result for the RVPN test site on June 1. If the spectral signature of the common test site surface is unknown and the spectral band difference effect is 5 percent.   85  .6740 0.891 14.3 percent in spectral bands 3 and 7.9050 1.7 percent overall.8 percent for bands 1-4 and 3.52 Spe ctral Ban d Prelaunch Calibration Gi5 (CPUR) Difference Relative to Prelaunch Gi5 (CPUR) 1 2 3 4 5 7 1.555 0.020 1. 1998). The average difference between the cross-calibration and vicarious calibration results is 1. the minimum and maximum differences being 1.081 7. These results are presented in Table 8. Error bars are not shown in Figure 12.111 8.944 14.243 0.8953 1.3% 1999 NIOB ETM+ Cross Calibrati on Gi5 (CPUR) 1.3% -2.2% -1.9% -9. then the overall uncertainty approaches ± 6 percent.786 1.6561 0.082 7.7% 1. respectively. but uncertainties in the vicarious calibration results are reported to be ± 5 percent in the visible and nearinfrared bands and approximately 50 percent greater in the short-wave infrared bands (Thome et al.980 7.261 0.7% 21% 1999 RVPN ETM+ Cross Calibratio n Gi5 (CPUR) 1..9050 1.52 1999 UAZ Vicario us Calibra tion Gi5 (CPUR) 1.6270 0.1 and 9. 1997). say.1% -9.681 16. The results in the figure are consistent with the possibility that there were no major changes in the Landsat-5 TM responsivity in any of the spectral bands between 1994 and 1999.875 14.8% 1.944 14.097 13.082 7.082 7.4% -1.8939 1. Table 8. 1999 has kindly been provided by UAZ in advance of publication. it should be noted that there is an additional 5 percent peak-to-peak variation in the general responsivity trend. likely due to the periodic build up of ice on the window in front of the cold focal plane (Markham et al. An evaluation of the tandem cross-calibration would be incomplete without a comparison against results obtained from the two TM vicarious calibrations and to pre-launch responsivity coefficients for Landsat 5.7 together with the RVPN ETM+/TM cross-calibration results.

Scott. Florence. Table 8.J. and Next-Generation Satellites V. Sabelhaus. J. Williams.A. Italy. Thompson.2 percent is found in spectral bands 1-3 and 5. the use of prelaunch calibration coefficients for the visible bands would lead to significant errors in TOA radiances and any quantities derived from TOA radiances. 2000). S.7 Summary The work described here indicates that the tandem cross-calibration approach provides a valuable "contemporary" calibration update for Landsat-5 TM solar reflective bands based on the excellent radiometric performance of Landsat-7 ETM+. Proceedings of SPIE Conference 3870. Irons. Arvidson. J.T. J. the SDSU result for ETM+ band 7 responsivity is approximately 6 percent higher than the nominal ETM+ value. whereas the differences are only a few percent or less in spectral bands 4. including surface reflectances retrieved from the TM imagery via atmospheric correction for example. however. D.J. 2. S. Table 8.. 1999.F. Europto SPIE Conference on Sensors.. J. Systems. and Storey..5 for these bands.   86  ..C. SDSU also carried out a vicarious calibration for the Landsat-7 ETM+ at Niobrara on the same day and the ETM+ band 4 results is approximately 10 percent lower than the nominal ETM+ responsivity provided by the Landsat Project Science Office.. J. T. 8. J.4. Agreement to within 3.L. and 7. R.. P. Markham. Similarly. Differences ranging from -11 to -20 percent exist for spectral bands 1. 5. Dolan. are not unexpected. REFERENCES Barker. Rapp.L.R. although unsatisfactory.4. Users. should use the Mi slopes derived from the RVPN tandem pair for updating the radiometry for Landsat 5 level 0 data.. Clearly.7 presents a comparison of the TM responsivity results from the cross-calibration work and the SDSU vicarious calibration... pp. Kane. when the tandem data were acquired (Black et al.South Dakota State University (SDSU) has kindly provided in advance of publication vicarious calibration results for Landsat-5 TM obtained at the NIOB test site on June 2. This is a possible explanation for the difference between the tandem cross-calibration and SDSU results for TM band 4 (-9. 2000. B.J. Initial trials of the approach with two different tandem image pairs yielded repeatable results for TM responsivity coefficients. which could explain part of the 21-percent difference found between the TM band 7 results.3 percent).7 also compares the 1999 TM calibration update to prelaunch responsivity coefficients for the six solar reflective spectral bands... On-orbit calibration updates for spectral bands 5 and 7 have always been characterized by greater uncertainties and so the results in Table 8.S.. and 3. Bolek.K. "Landsat-7 Mission and Early Results".. This work is ongoing and any improved cross-calibration updates will be made in the Handbook as necessary.L.

.R. P.. K.299-311..L. in press. G. 2.. The IAS provides the capability to routinely perform four types of geometric characterization to verify and monitor system geometric performance. "Radiometric Cross-Calibration of the Landsat-7 ETM+ and Landsat-5 TM Sensors Based on Tandem Data Sets". geodetic accuracy assessment to measure the absolute accuracy of Level 1Gs (systematic) corrected products.. 1997. Irish. Thome. Markham. and three types of geometric calibration to estimate improved values for key system geometric parameters. "Radiometric Calibration of Landsat". 63(7):853-858.. J. S. image-to-image registration assessment to measure and monitor multi-temporal image registration accuracy.C.5 Geometric Calibration Overview This chapter describes the geometric characterization and calibration activities that will be performed over the life of the Landsat 7 mission using the software tools developed as part of the Landsat 7 Image Assessment System .   87  .. Teillet. B. Fedosejevs. J. Remote Sensing of Environment. 3.M. Markham. and Biggar. Photogrammetric Engineering and Remote Sensing. band to band registration assessment to measure and monitor the relative alignment of the eight ETM+ spectral bands. and 4. Storey. P. R. J..L. The geometric characterizations include: 1.C. Barker. J. B. Slater. Barker... 8. geometric accuracy assessment to qualitatively and quantitatively evaluate residual internal geometric distortions within Level 1Gs images.

The second priority during the IOC period will be to verify and. After the IOC period.g. and 3. The results of these initial calibration activities will be used to verify that the system is performing within specifications and to create an initial post-launch release of the Calibration Parameter File. band-to-band. scan mirror calibration to measure and correct any systematic deviations in the ETM+ scan mirror along and across scan profiles. and image-to-image characterization capabilities. Refining the pre-launch sensor alignment knowledge is critical to ensure that the Level 1Gs product geodetic accuracy specification can be met. and to perform the initial sensor alignment calibration. if necessary. sensor alignment calibration to provide improved knowledge of the geometric relationship between the ETM+ optical axis and the Landsat 7 attitude control reference system.   88  . which meet the Landsat 7 geodetic accuracy requirements. ground control. ongoing calibration activities will monitor the stability of the Landsat 7 ETM+ system's geometric performance and attempt to identify and characterize any systematic variations in the system's geometric parameters. geometric. update the pre-launch focal plane (particularly band placement) and scan mirror profile calibrations. The most critical geometric calibration activities involve measuring and verifying the Landsat 7 ETM+ system performance during the Initial On-orbit Checkout (IOC) period using the geodetic.The geometric calibration capabilities provided by the IAS include: 1. temperature. This will include processing additional calibration scenes under a variety of acquisition conditions (e. orbital position.g. which can be used by the IAS or the Landsat 7 Level-1 Product Generation System (LPGS) to create Level 1G products. 2... Sufficient supporting data sets (e. terrain data) to perform these characterization and calibration activities must be available at launch. Techniques for measuring and estimating improved values for individual detector locations and delays are being researched and may be added to the IAS as a post-launch capability. ETM+ time on) to measure the system's geometric performance as a function of time. focal plane calibration to measure and provide improved estimates of the eight band center locations on the two ETM+ focal planes relative to the ETM+ optical axis. and location.

the Landsat 7 definitive ephemeris is used for geometrically correcting ETM+ data. Definitive Ephemeris If available. which measure a combination of orbit. As the Kalman processes additional precision correction solutions filter. In the context of Landsat 7. reducing the uncertainty due to orbital errors. this decision is based on the alignment bias covariance estimates generated by the Kalman filter. This calibration will use discrete ground control points in a set of pre-defined calibration reference scenes. an IAS analyst will decide that the alignment knowledge has changed enough to warrant generating an updated sensor alignment matrix for inclusion in the Calibration Parameter File. An ephemeris is a set of data that provides the assigned places of a celestial body (including a manmade satellite) for regular intervals.8. The IAS is required to estimate this alignment to an accuracy of 24 arc seconds (per axis) at least once per calendar quarter. The position and velocity information are used during product generation. attitude. Multiple precision correction solutions will be integrated using a Kalman filter algorithm to estimate the best-fit systematic alignment bias. Periodically. ephemeris data shows the position and velocity of the spacecraft at the time imagery is collected.   89  . a new alignment matrix is generated whenever a new version of the CPF was scheduled for release. A new set of CPF parameters are generated as soon as the bias estimate standard deviation move below the 24-arc second alignment accuracy requirement threshold. The primary challenge in alignment calibration is the need to estimate the underlying alignment trend (assumed initially to be a bias) from a series of precision correction solutions. During normal operations. The GSFC Flight Dynamics Facility (FDF) upon request provides this precise ephemeris from the IAS. Definitive ephemeris substantially improves the positional accuracy of the Level 1G product over predicted ephemeris. Initially. the filter's estimates of the alignment biases will improve. and alignment errors. Landsat 7 will have more accurate (estimated to be in the 10-50 meter range versus 133 meter accuracy for the ephemeris downlinked in the Payload Correction Data) post-pass definitive ephemeris data available for the alignment calibration test scenes.6 Sensor Alignment Calibration The goal of the sensor alignment calibration is to improve the in-flight knowledge of the relationship between the ETM+ instrument and the Landsat 7 navigation base reference.

7 Scan Mirror Calibration The behavior of the ETM+ scan mirror is measured and. Any significant deviations detected will be folded back into the Calibration Parameter File as updates to the mirror profile polynomial coefficients. if necessary. calibrated using the IAS scan mirror calibration capability. The . Geometric accuracy on the order of 30-50 (1 sigma) meters. SLC deviations will be indistinguishable from scan mirror deviations so we have chosen to model the deviations as part of the scan mirror motion. This process compares a terrain corrected image to a high accuracy reference image constructed from a higher resolution source. 8. The support data used to construct the terrain corrected image is used to generate test points which can be related to a particular time within a particular forward or reverse scan. Detecting systematic deviations which can be attributed to mirror motion requires reference points which can be uniquely associated with individual forward and reverse ETM+ scans and which provide a good distribution of observations as a function of scan angle. The Flight Operations Team processes this information to produce a refined or "definitive" ephemeris that shows the position and velocity of Landsat 7 in one-minute intervals. to detect systematic deviations of the scan mirror motion from its nominal profile. Tracking data are used to compute the actual spacecraft position and velocity for the last 61 hours and to predict these values for the next 72 hours. Engineers with the Landsat Program have completed a predicted versus definitive ephemeris analysis. The predicted ephemeris data are uploaded to the spacecraft daily. The current approach to acquiring such a control reference is to use spatially accurate reference imagery for one or more calibration areas. in fact. In practice. if needed. This information comes from the three US operated ground-receiving stations and is augmented by similar data from NASA's Tracking and Data Relay Satellites. The scan mirror calibration procedure will compare a precision and terrain corrected ETM+   90  .The Landsat 7 Mission Operations Center receives tracking data on a daily basis that shows the position and velocity of the Landsat 7 spacecraft. Level 1G products produced after March 29. Scan mirror calibration applies to both the along and across scan directions so it will detect and compensate for Scan Line Corrector (SLC) deviations as well. On-board software interpolates from this data to generate the positional information contained in the Payload Correction Data (PCD).MTL field "ephemeris_type" in the product metadata files identifies whether a product was created with definitive or predicted ephemeris. Comparisons to ground control points demonstrate the definitive ephemeris is. it will be possible to estimate corrections to the pre-launch scan mirror profiles. excluding terrain effects. Daily definitive ephemeris profiles have been archived since June 29. 2000 use definitive ephemeris if available. can be achieved when the definitive ephemeris is used to process the data. reliably more accurate than the predicted ephemeris. By comparing these test points to the reference image and analyzing the measured deviations as a function of scan direction and scan time. 1999 and are available for downloading.

The CPF is also sent to international ground stations via the Landsat 7 Mission Operations Center.8 Focal Plane Calibration The focal plane calibration operations involve measuring the alignment of the eight ETM+ bands to ensure that band registration accuracy meets the 0. detector placement calibration techniques are still under investigation and at this time are not a part of the focal plan calibration procedure. This is achieved by assessing new imagery on a daily basis. performing both radiometric and geometric calibration when needed. Landsat 7 ETM+ images of focal plane calibration test sites will be used to measure and calibrate the internal alignment of the detectors on the two ETM+ focal planes. 9. These points. This involves constructing an array of points in the ETM+ scan geometry. with known scan number and time in scan coordinates will be correlated with the reference image to measure the (sub-pixel) residual distortion. If the band-to-band comparisons detect any uncompensated misalignment the band placement calibration procedure will be used to update the band center location parameters in the Calibration Parameter File accordingly. The distortion patterns from many scans will be analyzed to detect systematic deviations from the pre-launch forward and reverse scan mirror profiles. Although it is anticipated that scenes with long linear features would be used to assess the alignment of individual detectors. Processing parameters are stored in the Calibration Parameter File (CPF). Band to band registration assessment requires scenes which contain significant high spatial frequency content that is common to all eight ETM+ bands. These test sites are selected based on image content rather than the availability of supporting data.panchromatic band image with the reference image constructed from USGS Digital Orthophoto Quadrangle (DOQ) data to detect within-scan mirror deviations.1 File Description The IAS is responsible for the sustained radiometric and geometric calibration of the Landsat 7 satellite and ETM+ and passing this knowledge to the user community. Detector to detector alignment will also be monitored to ensure that image discontinuities are not introduced by using incorrect detector locations and delays in the Level 1G image resampling process. which are mapped to the output terrain corrected product. which is stamped with applicability dates and sent to the LP-DAAC for storage and eventual bundling with outbound Level 0R products. and developing new processing parameters for creating level 1 products.   91  .28 pixel requirement as stated in the system specification. 8.

The first two parameters in the file. IAS to name the CPF uses the following file naming procedure: L7CPFyyyymmdd_yyyymmdd.9. Time Stamps The calibration parameter file is time stamped by IAS with an effective date range. The timed release of a new calibration parameter file must be maintained because of the UT1 time corrections and pole wander predictions included in the file. Irregular updates. The need for unique file sequence numbers becomes necessary as file contents change.1 Calibration Parameter File Updates IAS updates and distributes the calibration parameter file at least every 90 days. will not affect the regular 90day schedule. you can download and view all CPFs since launch. File Naming Conventions Through the course of the mission. The parameter file that accompanies an order has an effective date range that includes the acquisition date of the image ordered. designate the range and are of the form YYYY-MM-DD.1. At this web site. however. The Effective_End_Date for the most recent parameter file is its Effective_Date_Begin plus 90 days. The IAS maintains a CPF archive. These parameters span a 180-day interval time centered on the effective start date of the new IAS CPF. After this date the file is without applicable UT1 time predictions. a serial collection of CPFs is generated and sent to the LP-DAAC for coupling to 0R products. Effective_Date_Begin and Effective_Date_End.nn L7 = Constant for Landsat 7 CPF = 3-letter CPF designator yyyy = 4-digit effectivity starting year mm = 2-digit effectivity starting month dd = 2-digit effectivity starting day _ = Effectivity starting/ending date separator yyyy = 4-digit effectivity ending year mm = 2-digit effectivity ending month dd = 2-digit effectivity ending day nn = Sequence number for this file where:   92  . Updates will likely be more frequent during early orbit checkout and will also occur between the regular 90-day cycles whenever necessary. A distinct probability exists that a CPF will be replaced due to improved calibration parameters for a given period or perhaps due to file error.

The new file names and sequence numbers become: File 3 L7CPF19981128_19990225. New versions or updates are incremented by one. Two new CPFs are created that supersede the time period represented by file number three. and a new version of file number four is created. This reserve sequence number uniquely identifies the pre-launch CPF.02 L7CPF19981128_19990131.As an example.03 L7CPF19990201_19990225. The assigned file names would be as follows: File 1 File 2 File 3 File 4 L7CPF19980601_199808210. suppose a dead detector is discovered to have occurred on January 31.02 L7CPF19980830_19981127.03 L7CPF19990226_19990526. Using this example.01 It is worth noting the 00 sequence number assigned to the original CPF. version 2.02 L7CPF19990226_19990526.01 L7CPF19981128_19990225.01 L7CPF19990226_19990526. The effective_date_end assigned would be the original effective_date_end for the time period under consideration.00 L7CPF19980601_199808210. two CPFs with new names and effectivity date ranges are spawned for the time period under consideration.02 L7CPF19981128_19990225. 1999. Assuming this scenario.g.02 File 4   93  .01 L7CPF19980601_199808210. A post-problem CPF with a new file name would be created with an _effective_date_begin corresponding to the imaging date the problem occurred. The effective_date_begin remains unchanged. New versions of all other CPFs affected by the erroneous parameter also would be created. suppose four calibration files were created by the IAS on 90-day intervals and sent to the LP-DAAC during the first year of the mission. if a specific problem (e. This example assumes the effectivity dates do not change. however.01 L7CPF19980830_19981127. The effective_date_end for a new pre-problem CPF would change to the day before the problem occurred. The effectivity date range for a file can change.01 L7CPF19981128_19990225. detector outage) is discovered somewhere within the nominal 90-day effectivity range. Sequence numbers for subsequent time periods all begin with 01. Further suppose that the first file was updated twice and the second and third files were updated once.

or artifact removal parameters. The heading for each group is the actual ODL group name used in the CPF. To illustrate consider the first three parameters in the file: Effective_Date_Begin.2. 9. Group statements used to aid in file organization and enhance parsing granularity of parameter sets. These three parameters form their own group. Effective_Date_End.9. 2. A brief description of each group and their use various Landsat 7 systems follows.2 File Content The CPF supplies the radiometric and geometric correction parameters required during Level 1 processing to create superior products of uniform consistency across the Landsat 7 system. which is called FILE_ATTRIBUTES.2 File Structure All calibration parameters are stored as American Standard Code for Information Interchange (ASCII) text using the ODL syntax developed by JPL. radiometric parameters. They fall into one of three major categories: geometric parameters.1. ODL is a tagged keyword language developed to provide a human-readable data structure to encode data for simplified interchange.1 Geometry Parameters The geometric parameters are classified into 11 first tier groups. Attribute assignment statement used to assign values to parameters. and the CPF_File_Name. • Earth Constants • Orbit Parameters   94  .01 END_GROUP = FILE_ATTRIBUTES 9. The syntax employed for this collection of parameters in the CPF appears as: GROUP = FILE_ATTRIBUTES Effective_Date_Begin = 1999-02-26 Effective_Date_End = 1999-05-26 CPF_File_Name = L7CPF19990226_19990526. The body of the file is composed of two statement types: 1.

high gain signal noise. and a current gain for each of the two gain settings. low gain dynamic range quality. The prelaunch and post launch gains are based on the   95  . The heading for each group is the actual ODL group name used in the CPF. For each detector there is a prelaunch gain. • Detector Status The Detector Status parameters provide a five-element code that describes the current health status of each ETM+ detector.2. • Detector Gains Analysis of the SIS calibration transfer to the IC and output from the CRAM model used by IAS results in the Detector Gain parameter set. post launch gain. and high gain dynamic range quality. The five codes indicate detector status (live or dead).2 Radiometric Calibration Parameters The radiometric parameters are classified into 15 first tier groups. A brief description of each group and their use in various Landsat 7 systems or by user follows. low gain signal noise.• Scanner Parameters • Spacecraft Parameters • Mirror Parameters • Scan Line Corrector • Focal Plane Parameters • Attitude Parameters • Time Parameters • Transfer Function • UT1 Time Parameters 9.

reflective. A set of parameters exists for each of the three band groups . • MTF Compensation All image systems. and the length of useable IC data including the pulse.SIS calibration and remain static while the current gain is updated as a function of CRAM model improvement and detector responsivity over time.   96  . cause a blurring of the scene radiance field during image acquisition. This is a complex task and may be subject to anomalies. • Detector Biases B6 During level 1 processing band 6 biases are generally computed from the IC for the image being processed. These values are used by LPGS to convert 1G products to scaled 8bit values and by users for the reverse transformation. Five weighting parameters for both pixel and line directions were selected to best fit the optimal MTF response. including Landsat 7. Accurate characterization of this blurring is referred to as the modulation transfer function (MTF). The MTF compensation parameters are weighting functions for each band. They are used during radiometric correction for rapid retrieval of calibration pulse and shutter data. and thermal. • Bias Locations The bias location parameters refer to the IC data. These are baseline band 6 biases and are used during level 1 processing if the image specific IC-derived biases prove unreliable. Restoration processing compensates and corrects for systemic degradations to yield greater radiometric accuracy. There is an LMIN/LMAX pair per band for each of the gain modes. This parameter group is computed both prelaunch and at regular intervals over the life of the mission. panchromatic. The Detector Gain parameters are used to radiometrically correct ETM+ image data prior to LPS automatic cloud cover assessment (ACCA) algorithm and optionally by LPGS for as an alternative to computing gains on the fly from the IC data. These are the LMIN and LMAX values and are expressed in units of absolute spectral radiance. the length in pixels of the bias region. • Scaling Parameters The Scaling Parameter set consists of the lower and upper limit of the postcalibration dynamic range for each band in each gain state. They specify the starting pixel location for the bias (dark current restore). These are applied to the components of the piecewise cubic convolution kernel to generate the optimal MTF reconstruction kernel.

• Lamp Reference As explained above. The sensitivity temperature parameters describe the relationship between gain change and operating temperature for each detector and are used to adjust the gains derived from multi-calibration sources. • Lamp Radiance The lamp radiance parameters contain the actual radiance of the two IC lamps in three possible configurations (i. lamp 1 off . the radiance levels produced by the internal calibrator.lamp 2 off. The lamp reference parameters represent the instrument state at the time of calibration. The model that compensates for these effects requires as input 14 temperatures of the internal calibrator components and focal plane arrays. The reference temperatures are used to normalize the gains to a stable temperature. The lamp radiance parameters used to compute the gains used for converting raw ETM+ data to units of absolute radiance. A single reference temperature is calculated prelaunch and post launch for each band at both gain states. these temperatures are extracted from the PCD for the image being processed. For each detector there are 18 coefficients for both the low and high gain states.lamp 2 on). Most detectors show some dependence of responsivity with temperature. Post launch are determined using PASC and FASC calibration data. The reflective IC coefficients are used in the model that corrects for these effects. the IAS also performs a pre-launch calibration of the ETM+ and a post calibration using the combined radiometric model. Gains derived solely from IC data are not temperature adjusted. In general. and temperatures of the internal calibrator components and focal plane arrays. position in orbit. lamp 1 on .• Sensitivity Temperatures The temperature of the detectors on the primary focal plane of the ETM+ is not controlled and tends to warm up as the instrument operates.e. For each reflective band there are pre-launch. These parameters are instrument on time. Several parameters affecting instrument state are tracked and used for correcting this effect. or seen by the detectors vary as a function of instrument state. or seen by the detectors vary as function of instrument state. The cold focal plane is controlled but may operate at different set points. • Reference Temperatures The sensitivity temperature coefficients just described are used to adjust gains for varying focal plane temperatures. lamp 1 on .   97  . post-launch. and current values for the low and high gain settings. • Reflective IC Coefficients Radiance levels produced by the internal calibrator. Pre-launch values are established by transferring the SIS calibration to the IC lamps via the ETM+.lamp 2 on. However.

The temperatures are in a raw data format and require conversion to valid numbers prior to their use. 9. bias) values when processing the emissive IC data. They consist of a magnitude and time constant for each detector.3 Artifact Parameters The artifact parameters are classified into 9 first tier groups. The band 6 view coefficients capture this view and are used to adjust the contributing spectral radiances accordingly.• B6 View Coefficients The band 6 view coefficients are used in computing the actual shutter (i. Six conversion coefficients are supplied for each of the 28 different temperatures that accompany the PCD.e. The offset algorithm takes into account radiance of the shutter flag as well as contributions from other instrument components such as the scan mirror and scan line corrector. The heading for each group is the actual ODL group name used in the CPF. The scan mirror's contribution to the band 6 response must be calculated and accounted for. The emissive IC algorithm requires scan mirror temperature for computing band 6 gains and offsets. There are two coefficients for each lamp. A brief description of each group and their use in various Landsat 7 systems follows. The lamp coefficient parameters are used to linearly transform the raw counts to milli-amps.   98  . • Lamp Current Coefficients Included in the PCD are the currents for the two IC lamps. • Memory Effect Memory effect is a noise pattern commonly known as banding. The currents are in a raw data format and require conversion to engineering units (i. The thermistor coefficients parameters are used for this purpose.2. These are used in an inverse filtering operation to remove the memory effect artifact. Each band 6 detector has a different view of the contributing components. • Thermistor Coefficients Included in the PCD are a variety of ETM+ component temperatures. The memory effect parameters were derived by the IAS and are static. • B6 Temp Model Coefficients The Band 6 temperature coefficients are used to calculate the temperature of the scan mirror. It's observed as alternating lighter and darker horizontal scan-wide stripes.e. milliamps) prior to their use.

• Histogram Histogram analysis estimates the relative gains and biases for all detectors by characterizing the response behavior of individual detectors in a band relative to the other detectors in a band. These are included in the absolute gains and biases eventually applied. Two processing options are possible: linearly adjust the 1R data to match the means and standard deviations of each detector to a reference detector or to an average of all the detectors. and histogramming window size. The ghost pulse parameters identify the beginning and ending minor frames that bound this ghost image. The IAS employs these parameters for identifying and trending impulse noise in otherwise homogeneous data such as night scenes and FASC imagery. The impulse noise parameters specify a median filter width and an impulse noise threshold for each detector. the possibility of residual striping remains. saturation metrics. It appears in bands 5 and 7. • Impulse Noise Impulse noise within a digital signal manifests itself in a sample as a departure from the signal trend far in excess of that expected from random noise. Results are used to adjust the gains and biases applied during radiometric correction.• Ghost Pulse The ghost pulse is a faint secondary image of the internal calibrator lamp pulse. a normalization reference detector for each band. There is one striping parameter per band for each of the gain modes. Nonetheless. • Striping Striping is defined as residual detector-to-detector gain and offset variations within a band of radiometrically corrected (1R) data. The   99  . The histogram parameters control the algorithm by specifying detector noise. The coherent noise parameters consist of the number of noise components and a set of waveform characteristics that describe each component for each band. • Coherent Noise Coherent noise is a low-level periodic noise pattern that was found in all Landsat 5 imagery and characterized by the IAS for Landsat 7. The scan correlated shift parameters are derived by the IAS and are static. They consist of a bias magnitude for each detector and are used to compensate for the shift when it occurs. • Scan Correlated Shift Scan correlated shift is a sudden change in bias that occurs in all detectors simultaneously. The striping parameters are correction methodology flags. The 1R process is intended to remove detector-to-detector variations through the generation of relative gains and bias from histograms.

• Fill Patterns LPS uses two values to fill minor frames to distinguish missing or bad band data from good data.e. ratios. Although ACCA uses only bands 2 through 6. These may change during the mission and are therefore included in the CPF for LPS use. 9. Biases are reported in units of digital counts. • Detector Saturation In addition to normally observed saturation (i. or above 0 at the low end.waveform characteristics are the mean. The ACCA Biases parameter set is used in conjunction with the Detector Gains described above for converting raw digital numbers to units of absolute radiance.   100  . The ACCA algorithm uses a variety of threshold and band indices for cloud identification. The detector saturation parameters also identify these levels for each detector. the other band biases are included for completeness. There is one bias parameter per detector per band for each of the two gain modes. phase. sigma. LPS and possibly IGSs list the ACCA Threshold parameters in the CPF for use. The fill data is placed on a minor frame basis . The detector saturation parameters identify these levels for each detector. The cloud cover scores become searchable metadata and are used to filter out undesirable scenes from an archive interrogation. The alternating 0/255 fill pattern was selected to unambiguously flag artificial fill from reflectance values that naturally could occur. The two fill values used are zeros for odd detectors and 255s for even detectors. and magnitude.if data is missing from part of a minor frame the whole minor frame is filled. and indices to separate clouds from land. minimum. • ACCA Thresholds The LPS ACCA algorithm uses bands 2 through 6 in a combination of thresholds.2. • ACCA Biases The LPS automatic cloud cover recognition (ACCA) algorithm requires radiometrically corrected image data.4 ACCA Parameters Each scene processed by LPS undergoes automatic cloud cover assessment prior to being archived. and maximum for each component's frequency. An analog to digital converter may saturate below 255 counts at the high end. 255) two other types of detector saturation can occur. 0. Results are reported in metadata that eventually is used in client data searches. The analog electronic chain may saturate at a radiance corresponding to a level below 255 counts and above 0 counts on the low end.

000 band 3 1551. The parameter values listed in Table 9.000 band 4 1044.09 watts/(meter squared * ster * m2) K2 1282.• Solar Spectral Irradiances The LPS ACCA algorithm converts radiometrically corrected data to units of planetary reflectance prior to cloud filtering.700 band 7 82. which are listed in table 9. The transformation equation requires two calibration constants.07 band 8 1368. which reduces between-scene variability. Table 9. 7 and 8.000 • Thermal Constants ACCA converts Band 6 from spectral radiance to a more physically useful variable.2. There is one value for each band.000 band 2 1840.71 temperature degrees (Kelvin)   101  .2 ETM+ Thermal Constants Constant Value Units K1 666.1 Solar Spectral Irradiances (watts/m2/nm) band 1 1969. Table 9.1 are the mean solar spectral irradiances for bands 1 through 5. namely the effective at-satellite temperatures of the viewed Earth-atmosphere system.000 band 5 225. This involves normalizing image data for solar irradiance.

resulting in a precision geometry model. ETM+/Landsat 7 sensor/platform geometric model creation 4. Figure 10. geometric model precision correction using ground control 8. Payload Correction Data (PCD) processing 2.2 shows the process of creating a geometric correction grid and the application of that grid in the resampling process. 10. scan.10.1 describes the process involved in initialization and creation of the Landsat 7 geometry model. output space/input space correction grid generation 6. and deriving improved estimates of geometric calibration parameters such as the sensor to spacecraft alignment. terrain correction.3 describes the process of refining the Landsat 7 geometry model with ground control. longitude. sample) to geodetic object space (latitude.1 Overview The geometric algorithms used by Level 1 processing system at EDC were originally developed for the Landsat 7 Image Assessment System. image resampling 7. detector. The diagrams that follow describe the high-level processing flows for the IAS Level 1 processing algorithms. Figure 10. and height).2 Level 1 Algorithms The Level 1 processing algorithms include: 1. sensor line of sight generation and projection 5. Mirror Scan Correction Data (MSCD) processing 3. and knowledge of the Enhanced Thematic Mapper Plus instrument and Landsat 7 satellite geometry to relate locations in ETM+ image space (band. Figure 10. Figure 10. The overall purpose of the IAS geometric algorithms is to use Earth ellipsoid and terrain surface information in conjunction with spacecraft ephemeris and attitude data.   102  . characterizing the ETM+ absolute and relative geometric accuracy.4 again describes the creation of a geometric correction grid (this time precision). and resampling with terrain correction. These algorithms are used for purposes of creating accurate Level 1 output products.

Rectification and Resampling Flow Chart   103  .Figure 10.1 .Geometry Model Initialization Flow Chart Figure 10.2 .

and geometric characterization algorithms and finally an examination of the estimates of uncertainty (error analysis) associated with each of the algorithms. a discussion of the coordinate and time systems used by the algorithms and the relationships between them. Figure 10.3 . and solution procedures for the Level 1 processing.Precision Correction Flow Chart   104  . the mathematical development of. In this document are supporting theoretical concepts and mathematics of the IAS geometric algorithms. geometric calibration. a review of the Landsat 7 ETM+ viewing geometry.Detailed algorithm descriptions for each of main process boxes in these diagrams can be found in the Landsat 7 Image Assessment System (IAS) Geometric Algorithm Theoretical Basis Document (PDF).

4 . Ground control/reference images for geometric test sites .used in terrain correction and geometric calibration. Some of these algorithms also require additional ancillary input data sets. Digital elevation data for geometric test sites . geodetic accuracy assessment. Pre-launch ground calibration results including band/detector placement and   105  .used to minimize ephemeris error when performing sensor to spacecraft alignment calibration.3 Ancillary Data The Level 1 processing algorithms include: The Landsat 7 ETM+ geometric correction algorithms are applied to the wideband (image plus supporting PCD and MSCD) data contained in level 0R (raw) or 1R (radiometrically corrected) products. 4. and geometric calibration algorithms. Precise ephemeris from the Flight Dynamics Facility .Precision and Terrain Correction Flow Chart 10.Figure 10. 2.used in precision correction. 3. These include: 1.

gravity constants) and dynamic Earth model parameters (e. This involves shifting pixels by integer amounts to account for the alternating forward-reverse scanning pattern of the ETM+ sensor. which would leave the value added domain for commercial companies. UT1-UTC time corrections) . is 0R data .timing. Level 1) products for users. This is readily apparent when viewing a simulated 0R image..Simulated Level 0R Image data form that is marginally useful prior to radiometric and geometric correction. however.used in systematic model creation and incorporated into the Calibration Parameter File. A Landsat 7 0R product.g.g. the odd-even detector arrangement within each band.. The primary product for users and vendors seeking higher level processing. Earth parameters including: static Earth model parameters (e. The baseline program philosophy was to provide raw data only. LPS spatially reformats earth imagery and calibration data into Level 0R data. 11.an essentially raw Figure11. polar wander offsets. 5. to be used in the generation of the initial Calibration Parameter File.1 . A prevailing "wait and see" position by commercial vendors prompted NASA to add a systematic correction capability to ensure product availability. scan mirror profiles. does contain all the ancillary data required to perform these corrections including a calibration parameter file (CPF) generated by the Landsat 7 IAS. and attitude sensor data transfer functions (gyro and ADS).1 Level 0R Product Unlike earlier Landsat programs. ellipsoid axes.e. the Landsat 7 system was not originally designed to produce high level (i. and the   106  . however.

1 Product Size Figure 11.2. The path/row notation was originally employed to provide a standard designator for every nominal scene center and allow straightforward referencing without using longitude and latitude coordinates.Level 0R Product Alternatives Three options.1 kilometers. The standard WRS scene overlaps neighboring scenes along a path by 5% at the equator and has a width or cross track distance of 185 kilometers. • Standard Worldwide Reference System (WRS) Scene.detector offsets inherent to the focal plane array engineering design. the pixel shift parameters used are documented in the IAS CPF. 4. All LPS 0R corrections are reversible. depicted in Figure 11. exist when defining the size or spatial extent of a Landsat level 0R product ordered from the LP-DAAC. A path distance of 90 kilometers before and after a WRS center point defines the standard scene length of 180 km. The WRS indexes orbits (paths) and scene centers (rows) into a global grid system comprising 233 paths by 248 rows. Landsat 7 browse is framed according to WRS scenes. The standard WRS scene as defined for Landsats 4 and 5 was preserved as a product for Landsat 7. The distance between WRS center points along a path is 161. An ordered scene will cover the same geographic   107  . The LPS 0R output is HDF-EOS formatted and archived.1.2 . Vol. Details of the archival format can be found in the Landsat 7 System Wideband DFCB. 11. This length includes 20 scans of overlap with neighboring scenes.

The unique bands of ETM+ image data comprise nine of the data sets.9.once in low and the other in high gain mode. The data is laid out in a scan line sequential format in descending detector order (i. is at least one WRS scene in length. a product specific metadata file. • 11. In the latter case. all bands) scene-based 0R product it will have 23 distinct files. • Partial Subinterval A partial Landsat 7 subinterval can also be ordered. if you order a complete (i. and an HDF directory. The partial subinterval is dimensioned according to standard WRS scene width. Subintervals are caused by breaks in the wideband data stream due to communication dropouts and/or the inability of the spacecraft to transmit a complete observation (interval) within a single Landsat 7 contact period.extent observed in the browse with the following caveat. Partial scenes (less than 375 scans) may exist at the beginning or end of a subinterval due to the fact that imaging events do not always start and end on scene boundaries. An interval is a scheduled ETM+ image period along a WRS path. The largest possible subinterval is 35 scenes long. 1 . Band 6 is captured twice .2 Product Components A complete scene-sized 0R product consists of 19 data sets derived from the wideband telemetry. All image samples or pixels are 8 bits in size. A brief description of each follows.e. an IAS-generated calibration parameter file. a geolocation index generated by EOSDIS Core System (ECS).e. The smallest possible subinterval is a single ETM+ scene. and can be up to 10 scenes in length if ordered in 0R form or 3 scenes in length in 1G form. Under nominal satellite configuration the low gain form of band 6 will be present in format 1. detector 16 followed by detector 15 and so on for the 30 meter bands).   108  .1. • Subinterval. Earth Image Data . A partial subinterval can float or be positioned at any scan line starting point within a subinterval. Therefore. Partial subintervals are defined by either contiguous WRS locations or a bounding longitude/latitude rectangle. Browse and scene metadata for these occurrences accurately reflect their partial scene nature and geographic extent although partials are currently not offered due to complexities associated with level 1 processing. and may be from one to 90 scenes in length. all scan lines touched by the bounding rectangle are included in their entirety. A subinterval is a contiguous segment of raw wideband data received during a Landsat 7 contact period. Standard WRS scenes have 375 scans.

format 1 . 15.format 2.   109  . which actually applies to the previous scan. The left starting pixel offsets also apply to the IC data.format 2 A duplicate set of PCD is generated when format 2 is processed and is kept with the product in the event format 1 is lost or corrupted. During LPS processing image data is shifted in an extended buffer to account for predetermined detector and band shifts.the first half scan error.g. 16. PCD . and the scan line direction. A duplicate set of MSCD is generated when format 2 is processed and is kept with the product in the event format 1 MSCD is lost or corrupted.g. 13. 12. 11. Internal calibrator (IC) data . detector 16 followed by detector 15 and so on for the 30 meter bands).format 1. and possible bumper wear. Scan line offsets . The data is collected once per scan and structured in a band sequential format in descending detector order (e. Each logical record consists of 3 MSCD data values .10.format 2 . Also included in the MSCD file are scan based values such as time code. The scan line offsets represent the actual starting and ending pixel positions for valid (non-zero fill) Earth image data on a data line by data line basis for bands 1 through 6 low gain. MSCD . scan line length. Internal calibrator (IC) data . is used to compute deviations from nominal scan mirror profiles as measured on the ground and reported in the calibration parameter file. MSCD . PCD . gain status and processing errors encountered by LPS The MSCD is trimmed to fit the product ordered although one additional record is added to the file during the subsetting process due to the fact that scan error and direction information corresponds to the prior scan. This information. PCD for the entire subinterval is included with the 0R product regardless of the size of the data set ordered.IC data for format 1 consists of scan line ordered internal lamp and shutter data for bands 1-5 and blackbody radiance and shutter data for low gain band 6.format 1.format 1 The PCD for format 1 consists of attitude and ephemeris profiles as well high frequency jitter measurements. the second half scan error. 14.IC data for format 2 consists of scan ordered internal lamp and shutter data for bands 7 and 8 and blackbody radiance and shutter data for high gain band 6. The data is collected once per scan and structured in a band sequential format in descending detector order (e. detector 16 followed by detector 15 and so on for the 30 meter bands). A logical record of MSCD exists for each data scan present in the 0R product ordered.

This file. Scan line offsets . three metadata files. These are stamped with applicability dates and sent to the LPDAAC for storage and bundling with outbound 0R products. A user may order a subset of the available bands. A file containing all the pointers. and data quality for bands 1 through 6 low gain. the format 2 metadata does not include cloud cover assessment data or references to browse data products. scan line length. In all cases. During LPS processing image data is shifted in an extended buffer to account for predetermined detector and band shifts. Geolocation Index. Only the internal calibrator. in its entirety and original form. content. The geolocation index is also produced by ECS. 19. count in a 0R product. and possible bumper wear. accompanies the 0R product. and 8. 21. This table contains scene corner coordinates and their product-specific scan line numbers for bands at all three resolutions.high gain. Its purpose is provide for efficient subsetting of a 0R product. and the HDF directory. every product includes two PCD files. Calibration parameters.ECS. 7. During LPS format 1 processing metadata is generated that characterizes the subinterval's spatial extent. file size information. The left starting pixel offsets also apply to the IC data. however.17. Metadata . The scan line offsets represent the actual starting and ending pixel positions for valid (non-zero fill) earth image data on a data line by data line basis for bands 6 high gain through 8. A third metadata file generated by ECS during order processing. and data objects required to open and process the 0R product using the HDF library and interface routines. and earth image file counts are affected by a product possessing less than the full complement of bands. accompanies the 0R product. the CPF. 23. HDF Directory. 22. The IAS regularly updates the CPF to reflect changing radiometric and geometric parameters required for level 1 processing. This file. 18. This file contains product specific information such as corner coordinates and number of scans. Format 2 metadata is similar but not identical to format 1 metadata. two MSCD files. 20. Metadata . the scenerelated metadata is specific to bands 6 . in its entirety and original form. Also.format 1.format 2. The subinterval-related metadata contents are identical. scan line offset.   110  .format 2. Metadata . which will affect the actual file.

11.1.3 Product Format The product delivered to Landsat 7 data users is packaged in HDF - an open standard selected by NASA for Earth Observing System (EOS) data products. HDF is a selfdescribing format that allows an application to interpret the structure and contents of a file without outside information. HDF allows Landsat 0R products to be shared across different computer platforms without modification and is supported by a public domain software library consisting of access tools and various utilities. Product users are directed to the Landsat 7 0R Distribution Product Data Format Control Book, Volume 5 (PDF) for details regarding the HDF design used for the 0R product. Included are references to NCSA-authored documentation. New users should begin with Getting Started with HDF while the HDF User's Guide and HDF Reference Manual are excellent resources for the HDF programmer.

11.2 Level 1R Product
The 2008 single-product data policy changes at EROS made the Level 1R product option obsolete. The following paragraphs are only relevant only from a historical perspective. The Level 1R product is a radiometrically corrected 0R product. Radiometric correction is performed using either the CRAM gains in the CPF or gains computed on the fly from the IC data. The choice is available to a user when the product is ordered. The biases used are always calculated from the IC data. Image artifacts such as banding, striping, and scancorrelated shift are removed prior to radiometric correction. Radiometric corrections are not reversible. The 1R product geometry is identical to the input Level 0R data. During 1R product rendering image pixels are converted to units of absolute radiance using 32 bit floating-point calculations. Pixel values are then multiplied by 100 and converted to 16 bit integers prior to media output. Two digits of decimal precision are thus preserved. One merely divides each pixel value by 100 to convert the 1R image data back to radiance units. The 16-bit 1R product is twice the data volume of an alike 0R product. Note for band 6: A bias was found in the pre-launch calibration by a team of independent investigators post launch. This was corrected for in the LPGS processing system beginning Dec 20, 2000. For data processed before this, the 16 bit image radiances are 0.31 w/m2 ster um too high. Official Announcement In the fall of 2000, two Landsat 7 science team investigator groups discovered a band 6-calibration bias in Level 1 ETM+ data products emanating from the EDC (i.e.

 

111 

LPGS) at the EROS Data Center. This bias apparently results from limitations in the pre-launch calibration of the ETM+. The magnitude of the correction is estimated to be 0.31 W/(m^2 sr um) or about 3-4% radiance error at typical surface temperatures. This apparent systematic error in Band 6 radiance calibration translates into estimated temperatures derived from Landsat 7 ETM+ being about 3°C too high for typical Earth surface temperatures. To remedy the situation, several changes were made to the product generation software and CPF. The band 6 biases and instrument component view coefficients were changed in the October 1, 2000 release of the CPF. The calibration equation used by LPGS software was operationally updated on December 20, 2000. Users need to be aware of the impact these changes have on level 1 products. LPGS Level 1 products - pre 12/20/00 Users should subtract the bias value (0.31 W/(m^2 sr um)) from the radiances obtained from Level 1R and Level 1G data products generated since launch by LPGS for both high and low gain band 6 data. The changes made to the October 1, 2000 CPF effectively remove the temperature bias but only if the product generation software uses the changed CPF values. The LPGS software in place prior to December 20th does not. LPGS Level 1 products - post 12/20/00 Users can safely use temperatures derived from band 6 radiance values. No bias correction is necessary. Other Systems - pre 10/01/00 If the CPF gains and biases are used then the band 6 radiance values should be adjusted for the bias described in paragraph three above. Other Systems - post 10/01/00 If the CPF gains and biases are used then no adjustments are necessary after this date. Other systems that employ a band 6-calibration equation are outside NASA/USGS configuration and control. For accurate processing, please consult with your product provider.

 

112 

11.2.1 Product Size Two options existed for users when defining the size or spatial extent of a Landsat level 1R product ordered from the LP-DAAC.

• Standard Worldwide Reference System (WRS) Scene. The standard WRS scene, as defined above for the 0R product, could be ordered in 1R form. Partial scenes that may exist at the beginning and end of subintervals could also be ordered. • Partial Subinterval A partial subinterval could also be ordered in 1R form. Unlike the 0R product the 1R was limited to a maximum of 3 WRS scenes in size. The variably sized 1R product could float or be positioned at any scan line starting point within a subinterval. Alternatively, the product could be defined by up to three contiguous WRS locations. •
11.2.2 Product Components A complete scene-sized 1R product consists of 17 data sets derived from the wideband telemetry, an IAS-generated calibration parameter file, a product specific metadata file, a geolocation index generated by EOSDIS Core System (ECS), and an HDF directory. Therefore, if you order a complete (i.e. all bands) scene-based 0R product it will have 21 distinct files. There are two fewer data files than an alike 0R product due to the fact that the multiple PCD and MSCD files are merged into single consensus files. Please reference the 0R file product for individual file descriptions. A user could order a subset of the available bands, which affected the actual file, count in a 1R product. In all cases, however, every product included one consensus PCD file, one consensus MSCD files, three metadata files, the CPF, and the HDF directory. Only the internal calibrator, scan line offset, and earth image file counts were affected by a product possessing less than the full complement of bands. 11.2.3 Product Format The 1R product was delivered to users only in the HDF format. The HDF 0R and 1R formats are nearly identical. Exceptions include the united PCD and MSCD files and an enhanced product specific metadata file that reflects 1R correction characteristics. Please refer to the Landsat 7 0R Distribution Product Data Format Control Book, Volume 5 (PDF) for details regarding HDF specifics. Additional information unique to the 1R product can be found in the ESDIS Level 1 Product Generation system Output Files DFCB (PDF).

 

113 

Refined parameters from the CPF.g. satellite (e. and any archived scene. can be ordered through one of two EROS web portals (Product Ordering). curvature.   114  . The correction algorithms employed model the spacecraft and sensor using data generated by onboard computers during imaging events. which includes the attitude and ephemeris profiles. ground control points and a digital elevation model are also used to improve the overall geometric fidelity of the standard level-one terrain-corrected (L1T) product. The 2005 Global Land Survey is used as the source for GCPs while the primary terrain data is the Shuttle Radar Topographic Mission DEM. The end result is a geometrically rectified product free from distortions related to the sensor (e. jitter.3 Level 1G Product The 1G product available to users from EROS is a radiometrically and geometrically corrected Level 0R image. The WGS84 ellipsoid is employed as the Earth model for the Universal Transverse Mercator (UTM) coordinate transformation. view angle effects). regardless of cloud cover. Associated with the UTM projection is a unique set of projection parameters that flow from the USGS General Cartographic Transformation Package. Primary inputs are the PCD. and Earth (e. The L1T correction process utilizes both ground control points (GCP) and digital elevation models (DEM) to attain absolute geodetic accuracy.g. the definitive ephemeris (if available) and the MSCD. attitude deviations from nominal). Scenes that have a quality scores of 99 and less than 40 percent cloud cover are automatically processed. rotation. Geodetic accuracy of the L1T product depends on the accuracy of the GCPs and the resolution of the DEM used*.11.g. relief).

7) semi-major axis of the projection's ellipsoid semi-minor axis of the projection's ellipsoid latitude of the first standard parallel latitude of the second standard parallel latitude of the true scale The Landsat 7 level 1G product projection parameters are listed in Table 12.Table 11. Most projections do not require 13 parameters as evidenced by the empty table cells.1.   115  .e.2. Longitude of the projection's central meridian scale factor at the central meridian (transverse mercator) or center of projection (Oblique Mercator) false easting in the same units as the semi-major axis false northing in the same units as the semi-major axis latitude of first point on the projection's center line latitude of second point on the projection's center line longitude of first point on the projection's center line longitude of second point on the projection's center line longitude down below pole of map latitude of the projection origin path number for Landsat 7 using the World Reference System #2 number of the Landsat satellite (i.1 Landsat 7 1G Projection Parameters Required Parameters Projection Name Universal Transverse Mercator (UTM) Lambert Conformal Conic Polyconic Transverse Mercator Polar Stereographic Hotine Oblique Mercator A Hotine Oblique Mercator B Space Oblique Mercator B AziAng AzmthPt CentMer Factor FE FN Lat1 Lat2 Long1 Long2 LongPol OriginLat Path Satnum SMajor SMinor STDPR1 STDPR2 TrueScale 1 SMajor SMajor SMajor SMajor SMajor SMajor SMajor SMajor 2 SMinor SMinor SMinor SMinor SMinor SMinor SMinor SMinor Factor Factor Satnum AziAng Path Azmt hPt Factor 3 Zone STDPR1 STDPR 2 Cent Mer Cent Mer Cent Mer LonP ol OriginL at OriginL at OriginL at TrueSc ale OriginL at OriginL at FE FE FE FE FE FE FE FN FN FN FN FN FN FN Lon g1 Lat1 Long2 Lat 2 zero one one 4 5 6 7 8 9 10 11 12 13 Projection Parameter Definitions azimuth angle east of north for center projection line longitude of point on central meridian where AziAng occurs. Parameter definitions are listed in Table 12.

Parameter definitions are listed in Table 12.north up Output format . some lack GCPs and/or DEMs required for precision and terrain correction processing. and thermal Resampling kernel .   116  .1.3 . In these cases.2. Most projections do not require 13 parameters as evidenced by the empty table cells.GeoTIFF File transfer protocol (FTP) download only * While most scenes are processed to L1T.Level 1GT-systematic terrain (GCPs absent) or Level 1G-systematic (DEMs and GCPs absent). 30.WGS84 Image orientation . the best level of correction will be applied .   Figure 11. VNIR/SWIR.UTM with Polar Stereographic projection used for Antarctica scenes Ellipsoid .L1T* Pixel Size . 60 meters for panchromatic.15.Landsat 7 Resampling Kernel (CC) Versus Other Methods The Landsat 7 level 1G product projection parameters are listed in Table 12.During L1T processing the 0R image data undergoes two-dimensional resampling according to the following set of parameters: • • • • • • • • Correction level .Cubic Convolution (CC) Map projection .

11.3.1 Conversion to Radiance During 1G-product rendering image pixels are converted to units of absolute radiance using 32 bit floating-point calculations. Pixel values are then scaled to byte values prior to media output. The following equation is used to convert DN's in a 1G product back to radiance units: Lλ = Grescale * QCAL + Brescale which is also expressed as: Lλ = ((LMAXλ - LMINλ)/(QCALMAX-QCALMIN)) * (QCAL-QCALMIN) + LMINλ where: Lλ = Spectral Radiance at the sensor's aperture in watts/(meter squared * ster * µm) Grescale = Rescaled gain (the data product "gain" contained in the Level 1 product header or ancillary data record) in watts/(meter squared * ster * µm)/DN Brescale = Rescaled bias (the data product "offset" contained in the Level 1 product header or ancillary data record ) in watts/(meter squared * ster * µm) QCAL = the quantized calibrated pixel value in DN LMINλ = the spectral radiance that is scaled to QCALMIN in watts/(meter squared * ster * µm) LMAXλ = the spectral radiance that is scaled to QCALMAX in watts/(meter squared * ster * µm) QCALMIN = the minimum quantized calibrated pixel value (corresponding to LMINλ) in DN = 1 for LPGS products = 1 for NLAPS products processed after 4/4/2004 = 0 for NLAPS products processed before 4/5/2004 QCALMAX = the maximum quantized calibrated pixel value (corresponding to LMAXλ) in DN = 255

The LMINs and LMAXs are the spectral radiances for each band at digital numbers 0 or 1 and 255 (i.e. QCALMIN, QCALMAX), respectively. LPGS used 1 for QCALMIN while NLAPS used 0 for QCALMIN for data products processed before April 5, 2004. NLAPS from that date now uses 1 for the QCALMIN value. Other product differences exist as well. One LMIN/LMAX set exists for each gain state. These values will change slowly over time as the ETM+ detectors lose responsivity. Table 11.2 lists two sets of LMINs and LMAXs. The first set should be used for both LPGS and NLAPS 1G products created before July 1, 2000 and the second set for 1G products created after July 1, 2000. Please note the distinction between acquisition and processing dates. Use of the appropriate LMINs and

 

117 

LMAXs will ensure accurate conversion to radiance units. Note for band 6: A bias was found in the pre-launch calibration by a team of independent investigators post launch. This was corrected for in the LPGS processing system beginning Dec 20, 2000. For data processed before this, the image radiances given by the above transform are 0.31 w/m2 ster um too high. See the official announcement for more details. Note for the Multispectral Scanner (MSS), Thematic Mapper (TM), and Advanced Land Imager (ALI) sensors: the required radiometry constants are tabulated in this PDF file.

Band Number 1 2 3 4 5 6 7 8

Table 11.2 ETM+ Spectral Radiance Range watts/(meter squared * ster * µm) Processed Before July 1, Processed After July 1, 2000 2000 Low Gain High Gain Low Gain High Gain LM LM LM LMA LM LMAX LMAX LMAX IN IN IN X IN 297.5 194.3 293.7 191.6 6.2 6.2 6.2 6.2 303.4 202.4 300.9 196.5 6.0 6.0 6.4 6.4 235.5 158.6 234.4 152.9 4.5 4.5 5.0 5.0 235.0 157.5 241.1 157.4 4.5 4.5 5.1 5.1 47.70 31.76 47.57 31.06 1.0 1.0 1.0 1.0 0.0 17.04 3.2 12.65 0.0 17.04 3.2 12.65 10.80 0.3 16.60 0.3 10.932 0.3 16.54 0.3 5 5 5 5 244.00 158.40 243.1 158.3 5.0 5.0 4.7 4.7

 

118 

11.3.2 Radiance to Reflectance For relatively clear Landsat scenes, a reduction in between-scene variability can be achieved through a normalization for solar irradiance by converting spectral radiance, as calculated above, to planetary reflectance or albedo. This combined surface and atmospheric reflectance of the Earth is computed with the following formula: Where: = Unitless planetary reflectance = Spectral radiance at the sensor's aperture = Earth-Sun distance in astronomical units from an Excel file or interpolated from values listed in Table 11.4 = Mean solar exo-atmospheric irradiances from Table 11.3 = Solar zenith angle in degrees Table 11.3 ETM+ Solar Spectral Irradiances (generated using the Thuillier solar spectrum) watts/(meter squared * µm) 1997 1812 1533 1039 230.8 84.90 1362.

Band 1 2 3 4 5 7 8

Day of Year 1 15 32 46 60

Table 11.4 Earth-Sun Distance in Astronomical Units Day Day Day Day Distance of Distance of Distance of Distance of Year Year Year Year .98331 74 .99446 152 1.01403 227 1.01281 305 .98365 91 .99926 166 1.01577 242 1.00969 319 .98536 106 1.00353 182 1.01667 258 1.00566 335 .98774 121 1.00756 196 1.01646 274 1.00119 349 .99084 135 1.01087 213 1.01497 288 .99718 365

Distance .99253 .98916 .98608 .98426 .98333

 

119 

3.5. Partial scenes that may exist at the beginning and end of subintervals may be also be ordered. This is the effective at-satellite temperatures of the viewed Earth-atmosphere system under an assumption of unity emissivity and using pre-launch calibration constants listed in Table 11.5 ETM+ and TM Thermal Band Calibration Constants Constant 1. can be ordered in 1G form.K2 watts/(meter squared * ster * µm) Kelvin Landsat 7 666.   120  . Unlike the 0R product the 1G is limited to a maximum of 3 WRS scenes in size.09 1282. The variably sized 1G product can float or be positioned at any scan line starting point within a subinterval.4 Product Size The same two 1R options exist for users when defining the size or spatial extent of a Landsat level 1G product ordered from the LP-DAAC. The conversion formula is: Where: T K2 K1 L = = = = Effective at-satellite temperature in Kelvin Calibration constant 2 from Table 11.76 1260. The standard WRS scene.K1 Constant 2 .5 Calibration constant 1 from Table 11.56 11.5 Spectral radiance in watts/(meter squared * ster * µm) Table 11. • Partial Subinterval. Alternatively.3 Band 6 Conversion to Temperature ETM+ Band 6 imagery can also be converted from spectral radiance (as described above) to a more physically useful variable.11.71 Landsat 5 607. the product can be defined by up to three contiguous WRS locations.3. as defined above for the 0R product. • Standard Worldwide Reference System (WRS) Scene. A partial subinterval can also be ordered in 1G form.

A header file containing three records accompanies the image data.3. The 1R metadata file description can be found in the ESDIS Level 1 Product Generation System Output Files DFCB (PDF). external element style). The design employs external elements for the band files and metadata.   121  .exploit the NCSA HDF libraries for data access or process the data files directly using homegrown code. and geometric respectively.e. gains and biases can be found in the radiometric record while projection information and image coordinates are stored in the geometric record. which affects the actual file. a metadata file.3. The HDF directory and metadata files are always present regardless of bands ordered. A product with a full band complement has 11 files .6 Product Format The 1G product can be packaged into one of following user-specified output formats: • Hierarchical Data Format.11. Each band is self-contained in its own file (i. A user may order a subset of the available bands. Its structure is straightforwardly simple. The three records in order of appearance are labeled administrative. Please refer to the Landsat 7 0R Distribution Product Data Format Control Book. Volume 5 (PDF) for details regarding band file specifics. These are standalone files that are referenced via tags and pointers residing in an HDF directory.the HDF directory. radiometric. Sensor specific information is placed in the administrative record. The HDF format can be specified for any type of 1G product ordered from the LP-DAAC. 11.5 Product Components The 1G product ordered from the LP-DAAC consists of the corrected image files and descriptive metadata. count in a 1G product. The number of files comprising an HDF-formatted 1G product will vary according to the number of bands ordered. The Fast Format was originally developed by EOSAT as a means for quickly accessing Landsat 4 and 5 image data. The HDF packaging format used for the 0R and 1R products is also used for structuring the 1G. A single header file along with the image files constitutes the Fast product. External elements provide users with two processing options . and a separate file for each band. • Fast. All other ancillary files delivered with the 0R and 1R products are not included.

greyscale. & 60 meters) thus requiring a distinct header file for each. • GeoTIFF. compression size). For Fast-B and Fast-C all bands were resampled to a common grid cell size thus permitting a single header file. byte format. and scanned images between applications and computer platforms. GeoTIFF refers to TIFF files which have geographic (or cartographic) data embedded as tags within the TIFF file. The geographic data can then be used to position the image in the correct location and geometry on the screen of a geographic information display. The TIFF file consists of a number of label (tags). For simplicity's sake the grayscale model was implemented for the Landsat 7 GeoTIFF product. Under this implementation each ordered band is delivered as its own 8-bit greyscale GeoTIFF image. which describes certain properties of the file (such as gray levels. Geographic tagged image file format (GeoTIFF) is based on Adobe's TIFF . Today. Several differences are worth noting. In all likelihood each of the band groups for Landsat 7 will be resampled to a common resolution (i. the TIFF image file format is used to store and transfer digital satellite imagery. A full layout of the Fast L-7A Format can be found in the ESDIS Level 1 Product Generation system Output Files DFCB.e. elevation models. 30. After the initial tags comes the image data which may be interrupted by more descriptive tags. tiling. and extension to include geographic metadata. and output from digital cameras. 15. scanned aerial photos. No other files accompany the product. File names are now included in the administrative record which allows for direct file access. Baseline TIFF image types can be bi-level. TIFF is the only full-featured format in the public domain. capable of supporting compression. For detailed   122  . thermal and VNIR/SWIR band groups for Landsat 7. color table. logotypes. and full color (24 bit).a self-describing format developed to exchange raster images such as clipart. As a consequence Heritage Fast readers residing on user systems can be used for the Landsat 7 Fast formatted product.A derivative of the Fast Format (Fast-L7) used by EOSAT for Landsat (FAST-B) and Indian Remote Sensing products (Fast-C) was created for Landsat 7. A standard WRS scene possessing the full band complement would thus be comprised of nine separate GeoTIFF images or files. The Fast-L7 format supports all variations of the 1G product. All critical fields required for product ingest were left unchanged in the Fast L-7A Format. palette color. A separate header file now accompanies the panchromatic.

The center of a gap-present SLC-off data product is very similar in quality to previous Landsat 7 data. The problem was caused by failure of the Scan Line Corrector (SLC). This report includes input from scientists affiliated with the USGS. and these gaps will not be visible on the browse image preview when ordering SLC-off data. An instrument malfunction occurred onboard Landsat 7 on May 31. and the Landsat 7 Science Team. the scene's edges will contain alternating scan lines of missing data (Level 1G) or duplicated data (Level 0Rp or L1R). This product release includes all image data acquired by Landsat 7 in SLC-off mode from July 14. Landsat 7 ETM+ will therefore continue to acquire image data in the "SLC-off" mode.   123  . particularly within the central portion of any given scene. The precise location of the affected scan lines will vary from scene to scene. the USGS offers all archived Landsat scenes to the public at no charge. and the problem is permanent. 2008. Subsequent efforts to recover the SLC have not been successful. For GeoTIFF details. and Global Visualization L7 Image Browser. which compensates for the forward motion of the satellite. At the present time GeoTIFF format cannot be used for the Space Oblique Mercator and Oblique Mercator projections. 2003. 2003. However.information regarding the Landsat 7 GeoTIFF implementation please refer to the ESDIS Level 1 Product Generation system Output Files DFCB (PDF). As of November. Newly acquired Landsat 7 ETM+ SLC-off and Landsat 5 TM images with less than 40 percent cloud cover are automatically processed and made freely available for immediate download. 2003 to present. NASA. The first. The Landsat 7 Enhanced Thematic Mapper Plus (ETM+) is still capable of acquiring useful image data with the SLC turned off. excluding a 2-week interval from 9/3/03 to 9/17/03. EDC has recently released several Landsat 7 ETM+ SLC-off data products. The gap-present SLC-off data product is available as a single scene entity in Level 1G terrain corrected (L1T) form. please download the GeoTIFF Format Specification (PDF) or visit this web site. a gappresent product became available on October 22. A preliminary report regarding the utility of Landsat 7 SLC-off data is available in PDF form. Products projected into these reference systems must be formatted using HDF or Fast-L7. SLC-off data products can be searched and ordered via the Earth Explorer.

Middle image: scene after SLC anomaly. ephemeris data shows the position and velocity of the spacecraft at the time imagery is collected. An exampled of a gap-filled product is illustrated in Figure 11. An ephemeris is a set of data that provides the assigned places of a celestial body (including a manmade satellite) for regular intervals. Figure 11.7 Definitive Ephemeris If available. The position and velocity information are used during product generation. in conjunction with NASA. Bottom image: scene after SLC anomaly with data gaps filled. Various methodologies have been examined to fill the data gaps with observations acquired during prior or later than the target scene of interest. The two scenes are geometrically registered. 2004) is in 1G form and has the gap areas filled with Landsat 7 data acquired at a similar time of year and prior to the SLC failure.4. is continuing to research other methods of providing merged data products and will continue to provide information resulting from this work as it becomes available. the Landsat 7 definitive ephemeris is used for geometrically correcting ETM+ data.4 .3.The second product now being offered (as of May 10. which provides the best-expected radiance values for the missing data. In the context of Landsat 7. The USGS.Top image: pre-SLC anomaly. 11. and a histogram matching technique is applied to the fill pixels. Definitive ephemeris substantially improves the positional accuracy of the Level 1G product over predicted ephemeris. middle of image.   124  .

On-board software interpolates from this data to generate the positional information contained in the Payload Correction Data (PCD). Level 1G products produced after March 29. The LMIN corresponds to the radiance at the minimum quantized and calibrated data digital number (QCALMIN). can be achieved when the definitive ephemeris is used to process the data. Tracking data are used to compute the actual spacecraft position and velocity for the last 61 hours and to predict these values for the next 72 hours. in fact. 2000 use definitive ephemeris if available. 1999 and are available for downloading. excluding terrain effects. Engineers with the Landsat Program have completed a predicted versus definitive ephemeris analysis.99 of the pre-launch saturation radiance of the most sensitive detector in each band. The predicted ephemeris data are uploaded to the spacecraft daily. typically "255". which is typically "1" or "0" and LMAX corresponds to the radiance at the maximum quantized and calibrated data digital number (QCALMAX). Reflective bands: The LMIN's are set so that a "zero radiance" scene will still be on scale in the 8-bit output product.5 DN in high gain. Daily definitive ephemeris profiles have been archived since June 29. LMIN should result in "zero " radiance being about 5 DN in low gain and 7. Comparisons to ground control points demonstrate the definitive ephemeris is.   125  . 11. The Flight Operations Team processes this information to produce a refined or "definitive" ephemeris that shows the position and velocity of Landsat 7 in one-minute intervals.MTL field "ephemeris type" in the product metadata files identifies whether a product was created with definitive or predicted ephemeris. The LMAX's are set so that LMAX corresponds to slightly less than the saturation radiance of the most sensitive detector. reliably more accurate than the predicted ephemeris. This information comes from the three US operated ground-receiving stations and is augmented by similar data from NASA's Tracking and Data Relay Satellites. This is done so that in the output product all "pixels" saturate at the same radiance.3. even with sensor noise included. Geometric accuracy on the order of 30-50 (1 sigma) meters. Currently the LMAX is set to be 0. The .8 Radiometric Scaling Parameters for Landsat 7 ETM+ Level 1G Products The LMIN's and LMAX's are a representation of how the output Landsat ETM+ Level 1G data products are scaled in radiance units.The Landsat 7 Mission Operations Center receives tracking data on a daily basis that shows the position and velocity of the Landsat 7 spacecraft.

it was reported that the LPGS and NLAPS have different scaling procedures for scenes with gain state changes. Both systems always utilize the low gain LMAX value when scaling the radiance of a scene. This isn't the case. NLAPS trims the band offsets while LPGS does not.4 Level 1 Product Differences The two USGS Landsat 7 Level 1 product generation systems. is desired. Edge Trimming The two systems also treat the scenes scan line edges differently. LPGS and the National Landsat Archive Processing System (NLAPS). which were corrected. LPGS products are ordered through the EOS Data Gateway while Earth Explorer is used to order NLAPS products. The changes that have taken place to date have been mostly due to the adoption of "improved" pre-launch gains for the instrument that have. unless something changes drastically on the instrument. create products that are similar but not identical. This distinguishing characteristic (Figure 11. "increased" its sensitivity. The zero value is reserved for fill data. which is not expected. Scaling After radiometric correction the LPGS system scales the output values of the ETM+ 1G image data within the range of 1-255. The Landsat Project Science Office also detected a few errors in the original numbers.Normally. in effect. If the sensitivity decreases. If the sensitivity of the instrument increases. This change was made to assist in identifying the potential gaps for SLC-off products. 2004.5) can be used when system origin is unknown. prior to April 5. 11. NLAPS. Users should be aware of the following differences. scaled the 1G image data from 0-255 and also used 0 for fill data. Users must be cognizant of which system was used and the date of processing if a conversion back to radiance. 2004) ETM+ 1G data from 1-255 to match LPGS (the output scale for TM and MSS remains unchanged (0-255)) .   126  . Gain Change Scenes Earlier. as described in Chapter 11. the LMAX values can be increased which in turn increases the usable dynamic range of the product (this will not occur unless the change is large). there is no need to change the LMIN's or LMAX's. NLAPS now scales (post April 5. there is no need to change the LMIN and LMAX values.

NLAPS coordinates assigns coordinates to the upper left corner.5 where the upper left corner pixel is depicted for all three band resolutions. These differences are illustrated in figure 11.   127  . Map Projection Coordinates The two systems also use slightly different output projection pixel placement schemes. The NLAPS product also has the SampleFormat tag set to unsigned integer format.5 . GeoTIFF Header There is also a difference in the TIFF header for GeoTIFF formatted products. LPGS products have map coordinates that are pixel center based.Figure 11. The LPGS does not use the SampleFormat tag for its GeoTIFF formatted products. Figure 11.6 .The LPGS (left) does not trim scan edges like NLAPS (right). Both systems have the BitsPerSample tag set to 8.Upper left corner pixel alignment for LPGS versus NLAPS 1G products.

GXA Slewing Data anomalies that manifest as scan line offsets can occur when the Gimbaled X-band Antenna (GXA) is redirected from one ground station to another. Level 1 processing algorithms attempt to correct this problem but are not always successful. When not, LPGS estimates an end-of-line (EOL) and processes the damaged scans. The scans will be offset but available. NLAPS assumes an absent EOL means no data and will zero fill the affected scans.

12.1 Data Holdings
The primary receiving station for ETM+ data is the US Geological Survey's Center for Earth Resources Observation and Science (EROS) in Sioux Falls, South Dakota. Substantially cloud-free, land and coastal scenes are acquired by EROS through real-time downlink, and by playback from the on-board solid-state recorder. Approximately 200 scenes are downlinked or played back per day. Another 50-60 scenes arrive each day on tape from the northern latitude ground stations in Norway and Alaska. These scenes consist of Alaskan coverage as well as recorder dumps from foreign land masses. On April 21, 2008, the USGS announced plans to provide all archived Landsat scenes to the public at no charge. Newly acquired Landsat 7 ETM+ SLC-off and Landsat 5 TM images with less than 40 percent cloud cover are automatically processed and made freely available for immediate download. The downloadable inventory reaches back one year. Older scenes and imagery with more than 40 percent cloud cover can still be ordered, however, delivery takes days, not minutes. Only the L1T product is available. Previously offered USGS Landsat products with customer-defined projection options on various media are no longer available. There are a number of organizations and data centers that have made their ETM+ data holdings available online. This data is generally free or available at a nominal reproduction cost.

 

128 

Additional ETM+ Data Sources Source Geographic Area OhioView Ohio, Montana, SE Alaska, Turkey, Ethiopia U.S., Northern South America, Central Africa, Tropical Rain Forest Information Center SE Asia Chesapeake Bay, Canada, Caribbean, Northern Global Land Cover Facility South America, Philippines Sonoran desert region, including the US Arizona Regional Image Archive Southwest and northern Mexico Landsat.org Assorted scenes worldwide Landsat4u Southwestern U.S. W. M. Keck Earth Sciences and Mining Nevada Research Information Center EOS-WEBSTER Sparse U.S. and Russian collections Landsat 7 Data Sets Educational Image Downloads 12.1.1 Product Prices Landsat data is free and can be downloaded from the USGS's website GloVis and the Global Land Cover Facility's ESDI website. 12.1.2 Product Media Landsat products (both online and ordered) are distributed via FTP electronic transfer only. Once ordered scenes are processed, an email notification is sent to the customer with instructions for downloading.

12.2 Online Aids
The Landsat 7 satellite is part of NASA's Earth Observing System (EOS). The EOS Data and Information System component (EOSDIS) provides a structure for data management and user services for products derived from launched EOS satellite instruments, future missions and relevant NASA Earth science data for the foreseeable future. Within the EOSDIS framework, the Distributed Active Archive Centers (DAACs) are responsible for providing data and information services to support the customer community.

 

129 

These centers are responsible for data archival, product development and distribution along with user support. They are distinguished from one another by their data subject area. The LP DAAC is responsible for land processes data of which Landsat 7 is a part. The DAACs are linked by the Warehouse Inventory Search Tool (WIST) web portal which allows users to submit cross-discipline data (e.g. MODIS, ASTER) queries using spatial and temporal criteria, examine search results for relevancy using built-in tools, and submit orders via the EOS ClearingHOuse (ECHO) to the appropriate data providers. Landsat archive searching and downloading is now performed through the EarthExplorer and Global Visualization Viewer (GloVis) - two web portals developed by the EROS to replace the dated Global Land Information System (GLIS). EarthExplorer allows searches from Macs, PCs and Unix computers across multiple USGS maintained data sets. These data sets include Landsat 7, Landsats 1 - 5, AVHRR and aerial photography. GloVis simplifies the scene selection process via efficient retrospective examination of all acquisitions for a given WRS location. Two different product generation systems are used by EDC; the NASA-built Level 1 Product Generation System (LPGS) and the EDC-procured National Landsat Archive Production System (NLAPS). The two systems generate similar Landsat 7 Level 1 products but differences do exist. The user is allowed to choose the level 1 product rendering system. 12.2.1 Search Types An EarthExplorer search is performed by identifying the Landsat holding and then specifying a location via world map or place/address name. Additional criteria may include acceptable cloud cover, date range, and data type. A search commences and a results page is presented that lists all scenes meeting the search criteria. A set of links appears with each scene that allows one to examine the browse, download the scene, examine the scene's Earth footprint, and view the scene's metadata. Of particular metadata interest to users of ETM+ data are the cloud cover, data quality scores and the gain states for the individual bands, which are needed to convert the scaled digital numbers to radiance units. The GloVis portal provides a rapid way of examining the entire acquisition history for a specific WRS location. Once a collection type is selected from a pull-down menu, a user either enters a longitude-latitude coordinate or clicks on a world map to zero-in on the desired land area. A 3 by 3 Landsat browse grid appears with compass keys that allow sceneshifting navigation to the WRS of interest. A pull-down Map Layer menu allows for the overlay of land features such as major cities, rivers, roads, railways, and country boundaries. Search limits (e.g. cloud cover, date range) can be set using the Tools pull-down menu. Once   130 

5. band 8 covers more land area to the west than band 1 while band 1 covers more land area to the east. a blue margin exists on the left side of the browse due to band 3's more western position. and 6. When viewing a browse for geographic coverage it's important to keep the band staggering effect in mind. 5.2 ETM+ Scene-based Browse Figure 12. 1.1 . The browse image bands. 3. an on-demand order must be submitted.3 are placed in RGB color space. Band 5 reaches furthest to the east as evidenced by the red margin on the right side of the image. 7.ETM+ Sample Browse The land area seen in an ETM+ browse is not common to all bands. For example. For an image collected on a descending pass.2. The image planes are actually staggered according to the following band order arrangement: 8. the panchromatic band starts approximately 500 meters before band 5 but also ends 500 meters earlier. Knowledge of band offsets will prevent coverage   131  . 2. Though not as clear.a scene is selected and added to the order box it can be downloaded or ordered. 12. A downloadable message is splashed on the browse if the scene is currently online. 4. This band staggering effect can be seen in the browse image above.4. Otherwise.

12. This filtering operation works by examining each non-cloud pixel in the mask. 2000). Extreme differences are indicative of cloud signature corruption. Image pixels that fall below the new threshold qualify. While true in most cases. If 5 of the 8 neighbors are clouds then the pixel is reclassified as cloud. clouds are successfully separated from the desert terrain below. however. Omission errors are expected. descriptive statistics are calculated from the cloud category using band 6. Milstein). which preserves visual detail. A wavelets based technique. standard deviation. This two-pass approach differs from the single pass algorithm employed for Landsat-5. Twenty-six different filters are deployed for this purpose.3 Automated Cloud Cover Assessment The Landsat 7 Automated Cloud Cover Assessment (ACCA) algorithm recognizes clouds by passing through the scene data twice (Irish.surprises when examining a product for the first time. temperature inversions do occur. Three class categories result from pass one processing . a filter is applied to the mask to fill in cloud holes.clouds. Only the thermal band is examined during this pass to capture the remaining clouds. Unexpected cloud cover calculations will occur in these situations. and distribution skew. The final cloud cover percentage for the image is calculated based on the filtered cloud mask.3 demonstrate an example of the improved power the Landsat 7 ACCA algorithm has over the Landsat 5 algorithm. the pass two results are ignored and the cloud score reverts to the pass one result. and an ambiguous group that is revisited in pass two. After the two passes through the data. was used as an alternative to subsampling in creating the Landsat 7 browse (Ellison. The first pass through the data is designed to trap clouds and only clouds. The pass one goal is to develop a reliable cloud signature for use in pass two where the remaining clouds are identified. create algorithm havoc and must be minimized. During processing.   132  . Commission errors. When this occurs. non-clouds. The images in Figure 12. After processing. After pass one processing. These include mean temperature. New band 6 thresholds are developed from these statistics for use during pass two. The algorithm is based on the premise that clouds are colder than Earth surface features.2. In this case. the pass one and two cloud cover scores are compared. a cloud cover mask is created.

This desert area in Sudan (left.Figure 12. Here the clouds are captured while the rocks are recognized for what they are. row 31 scene in Table 5. 4. outputs from all 26 filters are presented for a path 28. PA S S . A .2 Landsat 5 and 7 Cloud Cover Mask Comparison . Additionally. It has been divided into five processes. Data preparation involves converting Bands-2 through -5 to planetary reflectance and Band-6 to apparent at-satellite radiance temperature. In the middle is the cloud mask generated !using the Landsat 7 algorithm.   133  . which parallel the five flow chart figures: Pass-1 Spectral Cloud Identification (filters 1-11) Band-6 Cloud Signature Development (filters 12-16) Pass-2 Thermal Band Cloud Separation (filters 17-20) Image-Based Cloud-Cover Assignments and Aggregation (filters 21-15) Nearest-Neighbor Cloud-Filling (filter 26) Most of the computer-intensive part of the processing is in Pass-1 and therefore most of the ACCA algorithm documentation below addresses the processes within Pass-1. This can be viewed in the cloud mask to the right.2. bands!5. 2) represents terrain where the heritage algorithm misclassified rocks as clouds due to their high !reflectance in band 5.4 ETM+ Automated Cloud Cover Assessment (ACCA) Algorithm Documentation of the ACCA Algorithm The operational Landsat-7 ETM+ Automated Cloud-Cover Assessment (ACCA) algorithm is described here.1 S P E C T R A L C L O U D ID E N T IF IC AT IO N The algorithm handles the cloud population in each scene uniquely by examining the raw uncalibrated Level-0R image data twice. 12.

q is the solar zenith angle. 1999). is related to Top-of-the-Atmosphere (TOA) spectral radiance Ll (in Watts/(m2 sr µm)) by Qi = Gi L*i + Q0i Where Gi is sensor responsivity (in DN per unit spectral radiance) for each detector in the band and Q0i is the average zero-radiance shutter background (in DN) from the Calibration Parameter File (CPF).Q0i = Gi L*i Thus. S. and d is the Earth-Sun distance in Astronomical Units. At-satellite temperature for Band-6 (T) is related to TOA spectral radiance (Ll) by T = K2 / 1n ((K1 / Ll) + 1) where T is the at-sensor temperature in Kelvin.For each spectral band. ground processing system at EDC. Sensor responsivity and the zero-radiance bias for each detector is maintained by the Image Assessment System (IAS) and recorded in the Calibration Parameter File (CPF) at the U. K2 is the calibration constant 1282. Bias-corrected image values are then given by ΔQi = Qi . TOA spectral radiances ( L*i ) are related to image data by L*i= ΔQi / Gi TOA planetary albedo or reflectance (r) for Bands-2 through -5 is related to TOA spectral radiance ( L*i ) in each band by r = L*i d2 / (E0i cos q) where E is the exo-atmospheric solar irradiance in each spectral band in Watts/(m2m). the 8-bit observed raw uncalibrated image quantized level Q. which are referenced in Table 1 and the Landsat Science Data Users Handbook (Irish. Radiometric detector normalization is applied in each spectral band. in units of digital number (DN).71 in degrees   134  .

Filter 2 Non-cloud/Ambiguous Discriminator. Hall discovered that NDSI values greater than .07 identifies potential low-reflectance clouds.07 are identified as non-clouds and are flagged as such in the cloud mask. cold clouds. Band3 Comparing each pixel entering this filter to a Band-3 threshold set at .g. A description of each filter. and undergoes modulation if snow or desert features are present in a scene.08. The profile is defined by the cloud populations mean. in Band-5. c. non-clouds. b. K1 is the calibration constant 666. The first pass through the data is designed to capture pixels that are unambiguously clouds and not something on the ground.. cirrostratus) were also eliminated. are passed to filter 3. and an ambiguous group of image pixels that are reexamined in pass two. Filter 1 Brightness Threshold Each Band-3 pixel in the scene is first compared to a brightness threshold. which is set at .7 qualify as potential clouds and are passed to filter 5.4 represent snow cover quite well. Five categories result from pass one processing warm clouds. Those falling below . The threshold was raised to . reflectance for clouds is high while snow is low. Pixels that fall between an NDSI range of -. Pixels that exceed this threshold are labeled as ambiguous and are re-examined in pass-2.25 and . The Band-6 temperature profile is formulated from the observed pass one cloud population if it exists. The reflectances of clouds and snow are similar in Band2.Kelvin. and L*i is spectral radiance from equation 3. However.7 to capture potential clouds of this type. The NDSI filter is expressed as: NDSI = (r2 r5) / (r2 + r5) This filter is designed to eliminate snow. Pixel values that exceed the Band-3 threshold. and skewness. Snow pixels that slip through are generally trapped later. variance. Pixels that fall below this threshold are identified as are passed to filter 2.   135  . Pixels outside this NDSI range are labeled as non-cloud and passed to Filter 4. Eight different filters are used to isolate clouds and to eliminate cloudless areas including problem land surface features such as snow and sand. This value was initially tried for ACCA to eliminate snow but clouds composed of ice crystals (e.09 in Watts/(m2 sr m). presented in the order implemented follows: a. 1995). desert. The pixels of clouds from pass one are used to develop a Band-6 thermal profile and thresholds for clouds for use in pass two where the remaining clouds are identified. Filter 3 Normalized Difference Snow Index The normalized difference snow index (NDSI) is used to detect snow (Hall et al.

Filter 6 Band 5/6 Composite The low values of the product of values are sensitive to the detection of clouds.0 is used. Pixels with a temperature a value less than 300K are passed to filter 6. reflectance for green leaves is high because very little energy is absorbed. In the near-infrared (Band-4). If a pixel value exceeds 300K. In the red region (Band-3).08 identifies potential low-reflectance clouds. The 4/3 ratio results in higher values for vegetation than for other scene features. Pixels that exceed this threshold are labeled ambiguous and are revisited in pass-2.d. It is particularly useful for eliminating cold land surface features that have low Band-5 reflectance such as snow and tundra. green leaves that are dead or dying absorb even less energy and are thus highly   136  . Band5 Comparing each pixel entering this filter to a Band-5 threshold set at . In the near-infrared (Band-4). including clouds. Filter 7 Non-cloud/Ambiguous Discriminator. Filter 4 Snow Threshold Knowledge of snow in a scene is important for pass two processing so a tally of snow pixels is retained.8 threshold qualify as snow and are recorded as non-cloud in the cloud mask. NDSI values above a . Sensitivity analysis demonstrated that a threshold setting of 225 works optimally. i. Pixels below this threshold are passed to filter 8 as possible clouds. Pixels that exceed this threshold are labeled as ambiguous and are re-examined in pass-2. h. Pixel values above this threshold are examined using filter 7. Filter 8 Band 4/3 Ratio for Growing Vegetation This filter eliminates highly reflective vegetation and is simply Band-4 reflectance divided by Band-3 reflectance. e. Pixels with ratios below this threshold are passed to filter 9. f. g. a realistic cloud temperature maximum. A threshold setting of 2. Those falling below . The Band 5/6 Composite is expressed as: Band 5/6 Composite = (1 r5) T This filter works because clouds have cold temperatures (< 300K) and are highly reflective in Band-5 and therefore have low values of (1 r5) T.08 are identified as non-clouds (probably water) and are flagged as such in the cloud mask. Filter 9 Band 4/2 Ratio for Senescing Vegetation This filter eliminates highly reflective senescing vegetation and is formed by dividing the Band-4 reflectance by the Band-2 reflectance. the chlorophyll in green leaves absorbs energy so reflectance is low. Filter 5 Temperature Threshold The Band-6 temperature (T) values are used to identify potential clouds. it is labeled as noncloud.

otherwise the cold and warm clouds are combined and treated as a single population.16248 works effectively. Filter 10 Band 4/5 Ratio for Soil This filter eliminates highly reflective rocks and sands in desert landscapes and is formed by dividing the Band-4 reflectance by the Band-5 reflectance. Rocks and sand tend to exhibit higher reflectance in Band-5 than in Band-4.6 C L O U D S IG N AT U R E D E V E L O P M E N T Pass-2 processing requires two new band 6 thresholds against which all ambiguous pixels are compared. Knowledge of desert pixels in a scene is important for pass-2 processing. If snow or desert rocks are present. In the green region (Band-2). the leaves absorb less energy because of chlorophyll loss and exhibit increased reflectivity.0 works effectively. the warm cloud class is eliminated. For each cloud pixel. Pixels above and below this threshold are classified as warm and cold clouds. A further separation into two classes is achieved by again using the Band 5/6 Composite filter. These two cloud classes are recorded in the cloud mask and used to develop two cloud signatures. Only the cold clouds are used if snow or desert soil is present. Snow was previously accounted for. The cloud thermal profile developed includes key statistics including the maximum cloud temperature. B A N D . Pixels with ratios that exceed this threshold are passed to filter 11. k.5. a desert pixel tally is retained. The 4/2 ratio values are higher for vegetation than other scene features including clouds. B . The desert indicator employed is simply the ratio of potential cloud pixels exiting and entering filter 10 compared against a threshold value of 0.reflective. mean. The at-launch setting was 2. Pixels that exceed this number are ambiguous and are revisited in pass-2. standard deviation and histogram skewness. If the remaining pixels are less than 50% the warm   137  . Filter 12 Snow and Desert Indicator An infrared/short-wave infrared ratio is used identify highly reflective rocks and sands in desert landscapes. whereas the reverse is true for clouds. one for the cold clouds and the other conjoined cloud classes.16 but was change in May of 2001(?) when the operational decision was made to Band-4 in low gain. Pixels with ratios below this threshold are passed to filter 10. A threshold setting of 2. A threshold setting of 1. Filter 11 Band 5/6 Composite for Warm and Cold Clouds All pixels reaching this filtering level are classified as clouds. respectively. Pixels that fall below this threshold are labeled ambiguous and are revisited in pass two. j. the Band 5/6 Composite is compared against a threshold setting of 210. a. Therefore. These thresholds are computed using the pass-1 cloud temperature statistics.

PA S S .0 e. The shift factor is set to 0. Desert and Mean Cloud Temperature Indicator Three conditions have to exist to continue pass-2 processing. The other is set high and is used to compute a less restrictive estimate of cloud cover. If a pixel falls below the high threshold it is labeled as a warm cloud.clouds are removed. If the histogram skewness is negative the cloud population is biased towards the left or colder tail of the distribution. Because of the upward bias in temperature an upward threshold adjustment is necessary. If any of these tests are not met processing passes to filter 22. C . All ambiguous pixels are tested against the two thresholds. The cold cloud scene percentage must be greater than 0.   138  . If all three test are met pass-2 processing continues at filter 15. Filter 13 Pass-1 Cloud-free Indicator The pass-1 cloud tally is compared against zero.5 and 83. The snow percentage for the scene is computed and compared against a threshold of 1%. The skewness becomes the skew factor if it is less than 1. If no clouds were tallied processing ends and the scene is declared cloud-free. If the scene is more than 1% snow the warm clouds are removed. b. otherwise it is set to 1.2 T H E R M A L B A N D C L O U D S E PA R AT IO N One of the thresholds is set low and is used to generate a conservative estimate of clouds in a scene. No skew factor is necessary because the thresholds are set at appropriate levels for identifying clouds that skew colder. d. It is relabeled cold cloud if it also falls below the lower threshold.4%. Filter 16 Temperature Histogram Positive Skewness If the histogram skewness is positive the cloud population is biased towards the right or warmer tail of the cloud temperature distribution.5 percentiles are the starting points for the two temperature thresholds and adjustments are made if necessary. the pass-1 cloud temperature mean must less than 295K and desert conditions must not exist. The histogram's 97.0. Filter 14 Pass-1 Cold Cloud.0. Filter 15 Temperature Histogram Negative Skewness Prior to computing the Band-6 thresholds a skew factor is computed from the skewness of the cloud temperature histogram. The thresholds are determined from the pass-1 cloud temperature histogram. c.

D . the 98.20. Cloud classes that qualify as legitimate are combined with the pass one clouds to form a single unified cloud class in the mask. Filter 19 Band-6 Warm Cloud Indicator Each ambiguous pixels Band-6 temperature is tested against the higher threshold and is labeled a warm cloud if it is lower. Filter 18 Band-6 Maximum Threshold A final check is made to see if the new upper threshold exceeds the histogram's 98. c.75 percentile (a threshold above or near the cloud temperature maximum is unwanted).B A S E D C L O U D .0 processing continues at filter 19. A3. Filter 17 Threshold Shift Deployment If the skew factor is positive. Both thresholds are adjusted by this amount.If a pixel temperature falls below the upper threshold. If so.C O V E R A S S IG N M E N T S A N D A G G R E G AT IO N After Band-6 is processed the scene percentages for the warm and cold cloud classes are computed.   139  . The presence of snow or desert features. IM A G E . upward adjustments are made to compensate for the warm cloud bias. If the skew factor is 0. If it is higher it is skipped and the next ambiguous pixel is likewise examined. Filter 20 Band-6 Cold Cloud Indicator Each pass-2 warm cloud is tested against the lower threshold and is re-labeled cold cloud if it is lower. and the magnitude of the two new cloud classes are used to accept or reject one or both classes.75 percentile becomes the new upper threshold and the lower threshold is adjusted by the amount of skewness compensation actually allowed. the cloud mask is flagged with a unique number that identifies a class of warmer clouds. Pass-2 then processing continues at filter 20. The threshold shift is the product of the skew factor and cloud temperature standard deviation. d. If the pixel temperature also falls below the lower threshold. b. The integrity of the two additional cloud classes is then appraised. the cloud-mask value is changed to a colder cloud-class identifier. a. The process continues until all ambiguous pixels are accounted for. Processing continues at filter 19 until all ambiguous pixels are accounted for.

If the scene is free from snow and desert soil the pass-1 cloud percentage is determined using both the cold and warm pass-1 clouds. The final scene cloud cover score is set to zero and processing ends. If any one condition is breached processing passes to filter 25. The pass-2 cold and warm cloud contribution cannot be more than 35%. Filter 23 Pass-1 Cloud Acceptance Indicator If the snow or desert conditions determined earlier exist the pass-1 cloud percentage is determined using the pass-1 cold clouds only. b. d. The mean of the pass-1 cold cloud population is again tested against the limit of 295K.4%. If no clouds are tallied. Additionally. If it is less the clouds are accepted as real but with less certainty. snow cannot be greater than 1% of the scene. the percentages of the scene represented by the pass-2 cold class and combined classes (cold and warm) are computed. Filter 24 Pass-2 Cold and Warm Cloud Acceptance The temperature means and maximums are computed for the pass-2 cold and warm cloud populations.   140  . Filter 21 Pass-2 Cloud-free Indicator The pass-2 cloud tally is compared against zero. Processing resumes at filter 26.a. If the mean is greater than 295K uncertainties exists. Processing resumes at filter 26. an event unlikely to happen. Filter 22 Pass-1 Cloud Temperature Mean This filter is used when the three pass-1 criteria are not met in filter 14. The pass-2 cold and warm clouds are united with the pass 1 if four conditions are met. If these four conditions are met all of the clouds identified in pass-2 are united with the pass-1 clouds and processing proceeds to filter 26. the final scene cloud cover score reported is the cold cloud % from Pass-1. c. the mean temperature for combined cloud population cannot be greater than 295K and the between the combined cloud maximum temperature and the upper threshold cannot be less than two degrees. If any of these tests are not met processing passes to filter 22. The final scene cloud cover score reported is the cold cloud % from Pass-1. The cold cloud scene percentage might be less than 0. the pass-1 cloud temperature might be greater than 295K and desert conditions may exist.

12. The formula for the combined score is: image score * 10 + PCD score The following paragraphs describe how the image quality and PCD quality scores are assigned. Filter 25 Pass-2 Cold Cloud Acceptance The pass-2 cold clouds are used if their contribution to scene cloud percentage is less than 25% and their mean temperature is less than 295 degrees. Filled pixels qualify as cloudy neighbors in subsequent tests. Filter 26 Threshold Shift Deployment Each non-cloud image pixel is examined and converted to cloud if at least 5 of its 8 neighbors are clouds.2. It is computed by dividing the total number of filled minor frames for a scene by 6313 (the nominal number of image data minor frames in a major frame for 30 meter bands).e. E . a. Uncertainty exists if either rule is breached and the pass-1 cold clouds are used in computing the final scene score. This operation boosts the cloud-cover content to more accurately reflect the amount of unusable image data in a scene.   141  . If these two conditions are satisfied the cold clouds are united with the pass-1 clouds and processing advances to filter 26. The second digit represents PCD quality and can range in value from 0 to 9. the cloud pixels in the mask are tabulated and a final cloud cover percentage score for the scene is computed. Afterwards. The first digit represents image data quality and can range in value from 0 to 9. This will give a number of equivalent bad scans.N E IG H B O R C L O U D .5 Algorithm for Calculation of Scene Quality A two-digit number that separates image and payload correction (PCD) quality is used by the LPS for Landsat 7.F IL L IN G A final step involves identifying and filling cloud mask holes. N E A R E S T. Image Quality Component The image quality digit is based on the number and distribution of bad scans or equivalent bad scans in a scene. Processing continues at filter 26.

Image Quality Component Image Quality no errors detected. What defines clustering versus scattering? It is proposed that bad scan lines are clustered if they occur within a grouping of 128 contiguous scans (approximately 1/3 of a scene). clustered <= 4 equivalent bad scans. There are approximately 7 PCD major frames for a standard WRS scene comprised of 375 scans.1 Scene Quality Score . scattered <= 64 equivalent bad scans. scattered <= 16 equivalent bad scans.   142  . scattered > 128 equivalent bad scans. Errors are characterized as scattered if they occur outside the bounds of contiguous PCD major frames. Like the image data. The image score is assigned according to the rules in Table 12.. clustered <= 64 equivalent bad scans. Errors are characterized as scattered if they occur outside the bounds of 128 contiguous scans. Table 12.1.The distribution of filled minor frames is characterized as being either clustered or scattered. A cluster of 128 bad scans will still yield a scene with a cluster 246 good scans which is almost 2/3 of a scene. A scattering of 128 band scans may make the entire image worthless.e. scattered (> 33% of the scene is bad) Score 9 8 7 6 5 4 3 2 1 0 PCD Quality Component The PCD quality digit is based on the number and distribution of filled PCD minor frames.384 bytes. Two minor frames correspond to a single scan while 256 minor frames (i. scattered <= 128 equivalent bad scans. it is proposed that bad PCD minor frames are clustered if they occur within a grouping of 2 contiguous PCD major frames (1/3 of a scene). 2 PCD major frames) correspond to 128 scans or approximately 1/3 of a scene. What defines clustering versus scattering? Each PCD minor frame has 16 jitter measurements and corresponds to 30 milliseconds or approximately 1/2 of a scan. The PCD score is assigned according to the rules in Table 12. clustered <= 16 equivalent bad scans. a perfect scene <= 4 equivalent bad scans. Clustering of filled PCD minor frames indicates that errors are localized whereas scattering indicates that numerous or all major frames may be affected. clustered <= 128 equivalent bad scans.2. Each PCD major frame consists of 128 minor frames or 16.

the lowest 00. scattered <= 256 bad minor frames. clustered <= 128 bad minor frames.Score 9 8 7 6 5 4 3 2 1 0 Table 12. clustered <= 32 bad minor frames. Table 12.3 IGS Landsat Data Archives The IGSs represent an evolving worldwide network of Landsat 7 data collection and product generation centers. 12. if not all. scattered > 256 bad minor frames. clustered <= 256 bad minor frames. Using this scoring system the highest possible rating for an image would be 99. an image with 16 filled scans that are scattered and with errorless PCD would have a 59 score whereas an image with intact image data and a 32 filled PCD minor frames that are scattered would receive a score of 95. scattered <= 128 bad minor frames. Most. scattered <= bad minor frames. For example. While foreign landmasses will be imaged for onboard storage and eventual downlink to US controlled ground stations the depth of ETM+ coverage at participating IGSs will be much deeper.   143  . Missing image data cannot be retrieved and thus impacts the user more severely than missing PCD. IGS data policy and product types may differ from those in the US.PCD Quality Component PCD Quality no PCD errors detected <= 8 bad minor frames.3 lists the participating IGSs and provides points of contact and web site addresses for online catalogs.2 Scene Quality Score . The score treats missing image data more critically than missing or filled PCD data. The score construct unambiguously alerts the user to image data deterioration. clustered <= 8 bad minor frames. The rationale is that PCD is less important because missing values can always be extrapolated or interpolated to enable level 1 processing. IGSs will capture every scene imaged by Landsat 7 within their respective acquisition circles. scattered (> 33% of the scene is bad) Scene Quality The score calculated using the methods described above is recorded in the scene level metadata under the keyword SCENE_QUALITY.

gov.de / German Aerospace Center (DLR) German Remote Sensing Data Center (DFD) D-82234 Oberpfaffenhofen GERMANY Mr.7/15/99 Online Catalog Dr.ca/ Canadian Centre for Remote Sensing 588 Booth Street Ottawa. Paul Trezise Manager Australia http://www.Schreier@dlr.54725999 Email: menes@conae.547. Km 40 P. Data Acquisition Division Canada http://www.Till@ccrs. ACT 2616 AUSTRALIA Phone: 61-2-62014100 Fax: 61-262516326 Email: P.O.au National Institute for Space Research Brazil http://www.net Mr.gov.it/ European Space Agency (ESA) via Galileo Galilei Casalla Postale 67 Frascati Rome 00046 ITALY Phone: 39-06-941-80610 Fax: 39-06941-80612 Email: Malbani@esrin. Box 28 Belconnen .gc.clirsen.ar/   144  .go v.Table 12.esrin.go v. Sao Paulo BRAZIL Phone: 55-12-560-9330 Fax: 55-12561-2088 Email: Serra@cptec.12/1999 Online Catalog Operational .trezise@auslig.ar Mr.html Recursos Naturales Por Sensores Remotos Edificio del Instituto Geografico Militar 4to Piso Apartado 17-08-8216 Quito ECUADOR Phone: 593-2-582447 Fax: 593-2581066 Email: ecotoecu@uio.esa.it Germany http://www.3 IGS Web Sites and Points of Contact Country WWW and Mailing Address Comision Nacional de Actividades Espaciales (CONAE) Ruta C 45 Km 8 Falda del Cármen (5187) Provincia de Córdoba ARGENTINA Phone: 54. Gunter Schreier Executive Director Phone: 49-8153-28-1375 Fax: 49-8153-28-1313 Email: Gunter.nz. Ontario K1A 0Y7 CANADA Phone: 613-947-1217 Fax: 613-9478201 Email: Susan. Mario Albani Executive Director European Space Agency http://odisseo.dlr.au/acres/index.7/15/99 Digital Catalog Operational . Paulo Roberto Martins Serra Head Operational .inpe.htm Australian Center for Remote Sensing (ACRES) P.7/15/99 Online Catalog Mr.c om/princpal.br Dr.br/ (INPE) Rodovia Presidente Dutra.auslig.e sa.satnet.gc.de Operational . Filemon Valencia Executive Director Online Catalog Operational .7/15/99 Digital Catalog Contact Lic.ccrs. Antonio MENESCARDI Station Manager Operational .nrcan. Box 01 Cachoeira Paulista.conae.ca Centro de Levantamientos Integrados de Ecuador http://www.7/15/99 Online Catalog Comments Argentina http://www.inpe.O.20090 Fax: 54. Susan Till Director.nrca n.

Takashi Nakazawa Associate Japan http://www. Torben N. CSIR P.go.FY 2000 Operational . Ike Marais Director South Africa http://www.soest.co.edu Dr.Mr. 70 Pekayon. Ohashi. HI 96822 USA Phone: 66-2-579-5218 Fax: 66-2-5613035 Email: Torben@hawaii.12/1999 Senior Engineer Operational . Bambang Tejasukmana Executive Remote Sensing Application Center Indonesia http://www.edu/ University of Hawaii at Manoa School of Ocean. 395 Pretoria 0001 SOUTH AFRICA Phone: 27-12-334-5055 or 334-5000 Fax: 27-12-334-5001 Email: Ike.jp/index_e.machi Hiki -Gun Saitama ken.go.id Mr.go.lapanrs.Huan -Xi Road Beijing 100086 PEOPLE'S REPUBLIC OF CHINA Phone: 86-10-625-87019 Fax: 86-10625-61215 Email: Xzpan@ne.t h/ National Research Council 196 Phahonyothin Road Ladyao . Chatuchak Bangkok 10900 THAILAND Phone: 808-956-3499 Fax: 808-9569399 Email: UserService@fc.nasda. Pasar Rebo Jakarta 13710 INDONESIA Telephone: 021-8717714 Fax: 021871771 Email: bankdata@lapanrs.c n/ Station (CRSSGS) Chinese Academy of Sciences No.go.rsgs. Pan Xizhe General Director Operational . 350-0393 JAPAN Phone: 81-492-98-1235 Fax: 81-49298-1001 Email: nakazawa.ac. Nares Chamboonroot Director of National Research Council (NRC) Thailand http://www.cn Mr. Numanoue .7/15/99 Remote Sensing Division Operational .th University of Hawaii at Manoa (UHM) United States http://www.ac.marais@csir. Earth Science and Technology 2525 Correa Road Honolulu.za Mr.za / CSIR Satellite Applications Centre (CSIR/SAC) Mikomtek .go .FY 2000 Online Catalog Online Catalog Director Online Catalog   145  .co.nrct. Nielsen Director Operational . 45 Bei -Shan.rsgs.html National Space Development Agency (NASDA) 1401.takashi@ nasda.O.jp China Remote Sensing Satellite Ground People's Republic of China http://www.c om/ Indonesian National Institute of Aeronautics and Space (LAPAN) JL Lapan No.sac. Hatoyama .FY2000 Online Catalog Online Catalog Prof.nrct.com or inderaja@pu.ha waii.

uncompensated attitude jitter. scan mirror instability)   146  . alignment knowledge).. without ground control. • Limited by spacecraft/instrument geometric model accuracy (e. excluding terrain effects.17 pixels (1 sigma) • Limited by focal plane alignment and stability Image-to-Image Registration • Geometrically corrected images from multiple dates shall be capable of being registered to an accuracy of 7. ephemeris.. Band-to-Band Registration • Geometrically corrected products shall have the multispectral bands registered to 0.1.3 meters (1 sigma) • Limited by high frequency distortions within images (e. attitude.13. They are: Absolute Geodetic Accuracy • Geometrically corrected products shall be accurate to 250 meters (1 sigma).g.1 Geometric Performance 13.1 Requirements The geometric performance of the ETM+ is judged against three key requirements placed on the Landsat 7 system.g.

Scenes are first radiometrically and geometrically corrected. Ephemeris data .1 .links image data to ephemeris and attitude 4. The clock performance is monitored by the Landsat 7 mission operations center and the ETM+ alignment is determined by ground processing calibration.orientation of payload to the spacecraft The ephemeris is estimated post-pass using tracking data. Any terrain effects are removed analytically in the comparison. On-board star trackers and gyros measure attitude. yaw) 3.Across and Along Track Mean Geodetic Accuracy! of ETM+ Products Since Launch 1. These are: Figure 13.1.13. Attitude data . ETM+ alignment .   147  . The product's geodetic accuracy depends on the accuracy of four data inputs. Control point locations are then measured on the processed imagery and compared to known ground locations. Spacecraft clock .spacecraft orientation (roll.2 Geodetic Accuracy Geodetic accuracy is monitored using calibration scenes containing ground control points.spacecraft position and velocity 2. pitch.

 The purpose of this assessment is to measure the relative alignment of the eight ETM+ spectral bands after processing to Level 1Gs for verification that the 0. Current trending reveals postcalibration (Figure 13. Multiple scenes with ground control are used to measure the systematic biases attributable to ETM+ alignment.3 .2) geodetic accuracy of systematic ETM+ products is approximately 80 meters (1 sigma). seven different calibration scenes were used. 13. If not. 1999.The ETM+ alignment is determined by measuring the orientation of the ETM+ payload relative to the L7 spacecraft attitude control system reference. Figure 13. A separate independent set of eighteen scenes covering same time span were used to verify the results. which is much better than 250 meter specification.17 pixel band-to-band registration requirement is being met. Periodic geodetic accuracy testing showed a slow build-up of along-track Figure 13.Across and Along Track Mean bias (Figure 13.1) after July. placed on the system. Twenty-four scenes acquired since July. A Geodetic Accuracy of ETM+ Products Since sensor alignment calibration update Launch was performed in June. The ETM+ alignment shows time-varying behavior and will continue to be monitored during the course of the mission. 1999 (~1 per cycle) were used to perform the calibration.3 Band-to-Band Registration Band-to-band registration assessment is performed periodically throughout the mission's life.1.1 . During the initial onorbit calibration during first ninety days. of 2000.Primary Focal Plane Assembly Across Scan Band !Offsets Since Launch   148  . the IAS remedies the band alignment by deriving new band center locations via band-toband registration calibration and updating the CPF.

Band calibration analysis showed a systematic shift in the band 5.17 pixel specification.10 pixels.  1999.08-0. The current calibration is time-dependent pending development of a temperature dependent model. The measured band registration accuracy was 0. 1999 although analysis revealed that band center estimates from individual scenes are still highly correlated with temperature telemetry.Cold Focal Plane Assembly Across Band revealed that registration Offsets! Since Launch between the primary and cold focal planes in the line direction increased to 0.   149  . The primary focal plane band centers (Figure 13.Band registration is monitored using desert scenes as they provide the best cross-band   correlation performance.08 pixels. 1999. The band center calibration was updated for data acquired after July. However.ETM+ Band Center Offsets Versus which is hotter than during the Specifications 90-day checkout. Measurements Figure 13.060.5) was well within 0. registration performance (Figure 13. Nonetheless. The band center locations measured prelaunch were evaluated using on-orbit data were updated using calibration scenes during the inorbit checkout period (first 90 days). the registration accuracy degraded after July.5 .4) more variable with a 3-4 microradian mean shift. 6. 7 locations after July.4 .3) are very stable but the cold focal plane band centers (Figure 13. The cold focal plane offsets coincide with change in ETM+ operating temperature range Figure 13.

  150  . the ETM+ images were compared against the DOQ reference images. The required ground control. Second. Coupled with the image registration analysis is the need to measure the ETM+ scan mirror  performance to ensure the pre-launch profile is correct.4 Multi-Temporal Image Registration The multi-temporal registration accuracy was determined using cloud-free scenes of the geometric calibration "super-sites". two ETM+ scenes were cross correlated. The term "geometric super-site" is used to describe those pre-selected WRS path-row locations for which ground control.13. digital terrain data. The geometric calibration super-site scenes were also used for this purpose. Also. Five cloud-free images of two separate calibration sites were used to measure registration accuracy. A minor adjustment to the cross scan profiles was made to model slightly non-linear scan line corrector behavior. and reference images were derived from digital orthophoto (DOQ) data. This provides a measure of image distortion that changes scene-to-scene although systematic calibration distortions may cancel out. The image registration assessment was performed in two ways. A DOQ reference image was constructed to provide full-width coverage of a Landsat 7 scene so that Figure 13.Pre and Post Scan Mirror Calibration measurements at all scan Registration !Accuracy angles could be obtained.1. an adjustment to the prelaunch scan angle monitor start/stop angles was made to improve image-to-image registration accuracy. The scan mirror calibration update was made to the CPF in the fourth quarter of 1999. First. This supporting data set makes it possible to produce precision and terrain corrected ETM+ images. This provides a measure of image distortion relative to an absolute reference. and reference imagery have been collected. and to compare them to accurately georegistered reference images.6 . No apparent problem was observed with the along scan mirror profile. terrain models. Mirror deviations as a function of scan angle were obtained by ??? These were crosscorrelating the ETM+ scene to the reference image. The one meter resolution DOQs were mosaicked and reduced to 15 meter resolution.

13. • All-scene average registration to DOQ: o meters along scan and 4. appears to be stable. this mirror behavior is similar to that observed in Landsat 5 TM. Specifically. The rate of growth (Figure 13. especially in the forward scans.1.7) of turnaround time.                   Figure 13.2 meters along scan and 1.8 meters across scan (1 sigma) • Two individual scenes fell outside specification versus DOQ.9 meters across scan (1 sigma) • All scenes were within specification. however. the active scan time is showing increased variability. This would effectively end the Landsat 7 mission.7 .ETM+ Scan Mirror Period Growth 151  . However.2 meters across scan (1 sigma) • ETM+ to ETM+ registration: o 3.Image registration accuracy (Figure 13. The image registration also improved using the updated scan mirror calibration parameters. probably caused by bumper wear.6) was measured before and after the scan mirror calibration. Analysis yielded the following results. lose synchronization with the calibration flag if scan time gets too long. the scan mirror could. Results revealed that the required image registration accuracy was achieved using the baseline prelaunch scan mirror calibration parameters. theoretically. • Registration to DOQ: o meters along scan and 4. Also.5 ETM+ Scan Mirror Stability Regular monitoring has revealed the ETM+ scan period is increasing with time due to  growth in turnaround time.7 meters across scan (1 sigma) • All-scene average ETM+ to ETM+ registration: o meters along scan and 1. The impact to ETM+ data is that scan gaps will gradually increase with time. However.

ETM+ GXA Data Anomaly and thus loses synchronization with calibration flag.8 . The scan mirror may lose synchronization or. The correlation was confirmed by a July.9) variation in scan line length.  Figure 13. The impact to ETM+ Data is dropped scans and calibration flag incursions (Figure 13. The scan mirror controller sees an "extra" timing pulse Figure 13.   152  . sometimes disturbs the ETM+ scan mirror. Most extreme examples exceed the maximum allowable scan length leading which leads to dropped scans. This occurs when GXA the stepper motor frequencies correspond to scan mirror harmonics. in extreme cases.8) at 705 km orbital altitude particularly in polar regions and within the South Atlantic anomaly.10 .ETM+ GXA Data Anomaly The gimbaled X-band antenna (GXA).Regions of High Electron Flux at 708 !Kilometers      Figure 13.10) into the Earth imaging area.9 . This phenomenon was not observed on earlier missions as pointable X-band antennae are new on Landsat 7. restart during imaging. when maneuvered in an across-track slew. The impact to ETM+ data is a wider than normal (Figure 13. Such  an occurrence is correlated with regions of high electron flux (Figure 13. 2000 solar flare which resulted in 14 anomalies in a single day. This phenomenon may have occurred on Landsat 5.

13. double-spanned (two 10m spans. Figure 13.7 Modulation Transfer Function (MTF) Characterization The MTF of the ETM+ is regularly measured for comparison to the pre-launch test results and for monitoring long-term instrument performance.   153  .12) over the Lake Ponchartrain bridge (Figure 13. middle of image. Landsat 7 Enhanced is still capable of acquiring useful image data with the SLC turned off. The bridge is long. Various interpolation and compositing schemes are currently being investigated to expand the coverage of useful data. with interpolation. Landsat 7 ETM+ will therefore continue to acquire image data in the "SLC-off" mode and has so since July 14.13) in Louisiana. straight.1.! Bottom image: Scene after SLC anomaly. The MTF characterization methodology involves the analysis of cloud free scenes (Figure 13. separated by 24.Top image: Pre-SLC anomaly scene. and the problem appears to be permanent. 13. 2003 the scan line corrector (SLC). and offers high signal contrast to the waters below. More information on the SLC-off anomaly can be found in Chapter 11.13.11 . Subsequent efforts to recover the SLC have not been successful. particularly within the central portion of any given scene. 2003.1. which compensates for the forward motion of the satellite. An interpolation example can be seen in Figure 13.4m) is approximately aligned with the Landsat ground track (<1°).! Middle image: Scene after SLC anomaly.6 Scan Line Corrector (SLC) Anomaly On May 31. failed at approximately 21:45.

Figure 13.ETM+ Images of Lake Ponchartrain Bridge   154  .12 .13 .ETM+ Image of the MTF Characterization Site Figure 13.

An inverse FFT is applied to convert the models to the space domain.The MTF characterization is performed on level 1R data. The root sum square (RSS) differences between the models and the data for both forward and reverse scans are then computed. Phase is determined by correlating each bridge segment to bridge templates (Figure 13. The forward and reverse scans are separated and classified according to bridge segment phase. The samples from the mean bins are interleaved and the end result is a 128 pixel oversampled bridge profile. The segments within this bin Bridge !Crossover range are then averaged resulting in 8 mean bin segments containing 16 samples each.   155  . Each segment consists of 16 pixels centered about a 3-point moving average. The image line numbers containing bridge crossovers (Figure 13.14) are identified and removed from the extracted image.15) templates shifted at 0. Adjusting the variable OTF parameters while keeping the detector model fixed also minimized the RSS differences. The range of 8 consecutive bins with the most Figure 13. The OTF is then used to compare the bridge models with the actual data. The forward and   reverse scan bridge models are multiplied by the OTF model in the frequency domain. gap between spans and the intensities of the two spans and background water as model parameters. The 1R data is in 16-bit scaled  radiance form with image artifacts removed. An optical transfer function (OTF) that borrows from previous Landsat 5 TM work was developed for ETM+. Each segment is assigned to a phase "bin" based on the offset of the best matching template. To minimize the RSS difference the bridge intensities were adjusted while keeping the span width and separation fixed.!Lake Ponchartrain segments is then identified. The OTF is a mathematical statement that describes the relationship between the input and output of an imaging system. A 256x2048 image window covering the bridge is manually identified and extracted.125 pixel increments from -1 to +1 pixels (17 total). It provides a complete measure of system performance in that it includes the phase relationship as well as the amplitude degradation of the bridge as the frequency changes. The data undergoes no geometric resampling. An analytical bridge model was constructed in the frequency domain using the bridge span width. The corresponding windows for the 30-meter bands are then extracted.14 . A bridge profile is then built by extracting the bridge segments from each data line. Numerous iterations yielded a final OTF model.

15 . South Dakota State University and the University of Arizona. a black body. which consists of two lamps.Lake Ponchartrain Bridge Crossover 13. Details on the devices can be found in Chapter 8.1 On-orbit Calibration Methods The three on-board ETM+ calibration devices are the Full Aperture Solar Calibrator (FASC) which is a white painted diffuser panel.   156  . a Partial Aperture Solar Calibrator (PASC) which is a set of optics that allow the ETM+ to image the sun through small holes and an Internal Calibrator (IC).2. The IAS has the role of monitoring the performance and calibration of the ETM+ instrument and providing updates to the calibration parameter file (CPF). Rochester Institute of Technology. The NASA/GSFC Landsat Project Science Office (LPSO) works with the IAS (located at EDC) in analyzing the calibration information and updating the algorithms used within the IAS. 13. a shutter and optics to transfer the energy from the sources to the focal plane. Additional funding from NASA supports vicarious radiometric calibration efforts at NASA/JPL. The results form the basis for updating the radiometric gain calibration parameters in the calibration parameter file.                           Figure 13.2 Radiometric Performance A significant improvement in the Landsat-7 system is the addition of the IAS Image as part of the ground processing system. The most recent results are presented below. Approximately every 6 months the scientists and analysts involved in characterizing ETM+ radiometric calibration meet and present their results.

. Two investigations are looking primarily at the reflective band calibrations: those of Helder and Thome (Thome. B.).ETM+ Band 1 Band Average High Figure 13. The lamps and black body provide the high radiance source for the bands. a calibration can be determined. J..and/or aircraft-based reflectance. Each of these investigations predicts the radiance at the sensor aperture using a combination of ground.7µm. et al.17 .3 Radiometric Performance Results  Figure 13. The investigations of Palluconi and Schott are concentrating on the thermal band (Schott.). K.. The IC provides a signal to the ETM+ detectors once each scan line as well as a view of the black shutter.2.The FASC is deployed in front of the ETM+ aperture approximately monthly.ETM+ Band 2 Band Gain! Calibration Results from On-board and High Gain! Calibration Results from Ground-Look !compared to Pre-launch values.            On-board and Ground-Look compared to Pre-launch values.J.L. Results will not be discussed for wavelengths less than 0. On-orbit performance of the calibrators can be found in a paper that covers the subject in great detail (Markham.   157  .2. 13. The shutter provides the dark reference for the reflective bands and a low temperature source for the thermal band. coupled with measured and/or modeled atmospheric parameters. Performance of the PASC has been anomalous and results are not included here. the IC has shown both short term and long-term instabilities. et al.16 . Based on the orientation of the panel relative to the sun and instrument and the pre-launch measured reflectance of the panel. radiance or temperature measurements. 13.).2 Ground Look Calibration (GLC) Methods There are four investigations evaluating the ETM+ radiometric calibration using GLC or vicarious methods. et al. At the short wavelengths.R.

The IC results have also been included. 13. e. The FASC trends are believed to be largely due to changes in the FASCs reflectance and not representative of the instrument. but the trends are small (less that 1. The combined calibration results for bands 1 -5.19.5%/year). Note. L. Figure 13. Note that in all bands.g. Figure 13. GLC and IC trends.17.ETM+ Band 5 Band Average High Gain! Calibration Results from On-board board and Ground-Look !compared to Pre-launch values. recognizing that part of the variability present is related to the IC itself. that there is some consistency between the FASC.ETM+ Band 7 Band Average High Gain Calibration Results from Onand Ground-Look !compared to Pre-launch values.21 . et al).22. The FASC results presented are based on the "best" portion of the FASC panel and have been adjusted based on an apparent 1° difference in the orientation of the panel from pre-launch measurements (Markhan. 7 and 8 are presented in Figures 13.18. however. 13. the vicarious results agree to within 5% of the FASC and pre-launch values and that the trends in the GLC results are not significant..20 . 13.ETM+ Band 3 Band Average High Gain !Calibration Results from On-board and Ground-Look !compared to Pre-launch values.   158  . respectively. The FASC results do show significant trends.21.20.ETM+ Band 4 Band Average High Gain Calibration Results from On-board and Ground-Look !compared to Pre-launch values. Figure 13. 13.16.19 .   Figure 13. B.18 . 13. 13. In each case the pre-launch calibration currently in the calibration parameter file is presented along with error bars representing <5%..

25 . This bias Figure 13. calibration parameter file on October 1. In band 6 the IC is the only on-board calibration device. Landsat Product Generation System (LPGS) product until December 20.24). there was no effect on the U.06%/year. the calibrated product radiance has a scatter of about 1% around the vicarious results (Figure 13.29. though significant.23.25). Although the coefficients were changed October 1. particularly relative to the Landsat 4 and 5 TM thermal bands.S.ETM+ Band 6 Band Average was originally measured as 0.24 . a calibration update will be performed. the calibration change may have been effective October 1. Figure 13.31 W/m2 sr Low Gain Mode Responsivity as mm and this correction was applied to the Determined by Use of the Internal Calibrator. After correction for the bias. 2000 when the software was revised. The ETM+ instrument also appears stable relative to the vicarious measurements.in band 7 all are increasing. The slope of the responsivity. Figure 13. Depending on how data are calibrated in non-US processing systems.ETM+ Band 6 In-situ Derived Top of Atmosphere (TOA) Radiances Versus Image Toe before bias correction for July 1999 to Present. Altering the shutter view coefficients in the calibration parameter file and changing the calibration equation implemented the 0. 2000. Updated measurements indicate the bias was closer to 0. If the consistency continues and the vicarious trends become significant.   159  . though a significant bias was detected (Figure 13.ETM+ Band 6 In-situ Derived Top of Atmosphere (TOA) Radiances Versus Image TOA after! bias correction for July 1999 to Present. This system is remarkably stable. The response to the IC over the life of the mission is shown in Figure 13.23 . 2000 or may have been later.31 correction. shows a change of less than 0. 2000.

timber harvest. very few comprehensive studies have been done to address changes in the Earth's temperate conifer forests. The practice of clearcutting sections of Washington's Olympic National Forest and other state forests in the Pacific Northwest was prevalent up until the late 1980s when changes in public policy caused logging to move from public to private land. researchers have only been able to monitor changes in specific locations with Landsat data due to its limited availability. Woodcock found that Landsat images could recognize areas where trees were dying due to lack of water. Woodcock and colleagues used Landsat to monitor how drought in the late 1980s and early 1990s affected forests in California's Sierra Nevada.1%/year in the thermal band. Temperate conifer forests lie at latitudes above tropical forests and below boreal forests and account for much of the forested area in the United States and Europe. Woodcock and colleagues plan to create a global monitoring system for temperate conifer forests.1 Monitoring Temperate Forests Curtis E.2. During the drought. While much attention is paid to deforestation in tropical rainforests. Boston University geographer Curtis E. The uncertainty in the calibration is estimated at <5% in the reflective and ~1% in the thermal regions. Woodcock . Previously. and humans all can destroy forests and their ecosystems. The monitoring system will measure the rates of destruction of conifer forests due to natural causes such as drought and fire and anthropogenic clearing due to harvest or development of forest lands.   160  . With the help of the frequent and comprehensive coverage of Landsat 7. watershed management. drought.Boston University curtis@bu. 14. and understanding the role of human activities on changes in regional climates. Understanding changes occurring in temperate conifer forests is important for understanding environmental issues including wildlife habitat protection. a factor making the trees more susceptible to disease and the forest more susceptible to fire. The monitoring system will also track the regrowth of forests and successional change in vegetation.4 Summary On-orbit results indicate that the Landsat-7 ETM+ absolute radiometric calibration is stable to better than 1.5%/year in the reflective bands and 0.13.edu Fire.

1 Using Landsat images of Washington's Olympic Peninsula (above). and will be used to identify large areas of significant changes in forest lands. Following up with the finer spatial resolution data from Landsat will allow determination of the type of changes and their geographic extent. Boston University researchers can keep track of what areas are being cut. and what areas of forest are regrowing. The square box in this 1986 image represents a square kilometer area within the Olympic National Forest.   161  . The MODIS instrument will fly aboard the Terra satellite set for launch in July 1999. Figure 14.The new system will work in conjunction with NASA EOS land cover change studies based on the EOS Moderate-resolution Imaging Spectroradiometer (MODIS).

3   162  .2 Figure 14.Figure 14.

there is evidence of regrowing vegetation in the clearcut patches. they create a database of areas on the volcano that are most prone to lava flow hazards. To understand the complex "plumbing" beneath active volcanic lave lakes and determine the amount of lava flowing from them. this portion lies within the Olympic National Park. Luke Flynn of the University of Hawaii has been using time series of Landsat images. which has been continually erupting since 1983. Flynn and Harris have also been working with Landsat data of active volcanic lava lakes around the world. Once they have compiled extensive observations of an individual volcano. they are studying long-term observations of eruptions in Mexico (Popocatepetl) and Guatemala (Santaguito and Pacaya). the forest is undisturbed. Flynn's colleague Andrew Harris is generating estimates of the amount of lava erupting onto the surface. Using similar Landsat data.The two sub-scenes above are from September.and one critical to many residents of Hawaii . Much of his work has focused on the persistence of volcanic eruptions at Hawaii's Kilauea volcano. In each sub-scene.is to map active lava flows and provide advance warning to public safety officials about these natural hazards. He plans to produce new Landsat 7 volcano (and wildfire) hazard maps for the State of Hawaii every 8 days.hawaii.2 Mapping Volcanic Surface Deposits Luke Flynn . 1987 (top) and September. Another objective of Flynn's research . In addition to their work in Hawaii. However. With Landsat observations of the heat emitted during eruptions. 1995 (bottom). Flynn can distinguish active lava flows from older flows that have already begun to cool. Much of the clearcutting occurred prior to 1984. Flynn plans to produce even higher resolution maps of active lava flows (15 m) with Landsat 7.edu. Flynn plans to collect Landsat 7 data at a ground station in Hawaii as the satellite passes over the state. After 1987.University of Hawaii flynn@pgd. Flynn and other volcanologists have been using remote-sensing data from the geostationary GOES satellite to monitor volcanic eruptions in remote areas in real time. The higher resolution of Landsat data (30 meters as compared to 4 kilometers for GOES) can produce maps of lava flows with pinpoint accuracy. according to Flynn. In the southern portion of the images. 14. there are clearcut patches (red) in the northern two thirds of the image. With these maps researchers can study the evolution of individual eruptions while they are taking place.   163  . Flynn produces maps of the leading edges of wildfires. With this data. new clearcuts are evident in images of this area through 1987.

Figure 14.                       164  .4 Landsat 7 and Land-based Views of Two Guatemalan Volcanoes.

html describes his coral reef mapping and classification efforts using space-based imagery. Topographic expression is vertically exaggerated two times.14.   165  .Mount Kilimanjaro Perspective View Volcanoes.4 Coral Reef Mapping Serge Andrefouet.6 MB) found at http://newlandsathandbook.3 Three Dimensional Land Surface Simulations This 3-D view was generated using topographic data from the Shuttle Radar Topography Mission (SRTM). a remote sensing specialist at the University of South Florida (USF).nasa.gsfc. This powerpoint presentation (41. Date acquired: February 200 (SRTM). a Landsat 7 satellite image. 2000 (Landsat 7)!   14. February 21. and a false sky rendering.gov/apps/prog_sect14_4. has been observing reefs around the world for the past five years. Figure 14.5 .

(2) The process of searching for and locking onto a received signal. it is the amount of rotation that a spacecraft undergoes per unit time. Angular measurement of the physical position of the optical axis with respect to the primary space vehicle reference axes.   166  . The process by which electromagnetic radiation (EMR) is assimilated and converted into other forms of energy. (2) The reflectivity of a body as compared to that of a perfectly diffusing surface at the same distance from the Sun. (1) The ratio of the amount of electromagnetic energy reflected by a surface to the amount of energy incident upon it. these processes involve no energy retention or transformation and are distinct from absorption. angle of drift. Abbreviated as A/D conversion.059 mrad/sec ((233 paths/cycle * 2*pi*1000 mrad/path) / (16 days/cycle * 86400 sec/day)). analog-to-digital conversion. the datum usually being mean sea level. Albedo may refer to the entire solar spectrum or merely to the visible portion. acquisition. The angle between the heading of the axis of a craft and its ground track. or scattering. however. diffraction. often expressed as a percentage. angular velocity. and normal to the incident radiation. Refers to point above the Earth's surface rather than those on it (elevation). absorption band. altitude. (1) Image captured by satellite sensor. Absorption takes place only on the EMR that enters a medium. ETM+ data undergo such a conversion prior to downlinking. Height above a datum. A substance that absorbs EMR may also be a medium of refraction. alignment data.Glossary A absorption. A range of wavelengths (or frequencies) of electromagnetic radiation that is assimilated by the atmosphere or other substance. The process of sampling continuous analog signals in order to convert them into a stream of digital values. Also called rotational velocity. albedo. For Landsat 7 it is equal to 1. primarily heat.

The radiance at the aperture of the sensor. Usually. band sequential. An interval in the electromagnetic spectrum defined by two wavelengths. Roll is the deviation from the vertical (the angle between the z-axis of the vehicle and the vertical axis. or angular rotation around the y-axis). bands designate the specific wavelength intervals at which images are acquired. spectral. x for the crosstrack axis. An opening that admits electromagnetic radiation to a detector or film. or wave numbers. Expressed in degrees. apogee.anomaly. Yaw is rotation around the z-axis. The arc of the horizon measured clockwise from the north point to the point referenced. perpendicular to the direction of flight. frequencies. at-aperture-radiance. attenuation. azimuth. ascending node. A format that arranges the data by band such that all of the data from band 1 followed by all of the data from band 2. and z for the vertical axis. The reduction in the intensity of radiation with distance from its source due to atmospheric absorption and/or scattering. A deviation from the norm. especially of a manmade satellite. not location. It does not include the inverse-square decrease of intensity of radiation with distance from the source. B background. The point in the orbit of a heavenly body. Azimuth indicates direction. attitude. With Landsat. The point at which the orbit of an earth satellite intersects the plane of the equator going from south to north. etc. Y is used for the axis that defines the direction of flight. Any effect in a sensor or other apparatus or system above which the phenomenon of interest must manifest itself before it can be observed. band. at which it is farthest from the Earth.   167  . Pitch is the angular rotation around the x-axis. The angular orientation of a spacecraft as determined by the relationship between its axes and some reference line or plane or some fixed system of axes. An example would be the lens diaphragm opening in a camera. aperture.

In Landsat parlance. or a geometric test pattern. In remote sensing. green assigned to band 2. this light is of a single wavelength. A color image produced by the combination of three individual monochrome images in which each is assigned a given color. it reflects or. The application of calibration data to restore measurements to their true values is called rectification. which is dependent on the wavelength of the light. the color seen is a pure spectral color. A high-order resampling technique is which the brightness value of a pixel in a corrected image is interpolated from the brightness values of the 16 nearest pixels around the location of the corrected pixel. absorbing all incident radiation. An error detection and correction scheme named after its inventors Bose.   168  . a temperature plate. if it existed. and emitting radiation at wavelengths. in the case of a luminescent body. and Hochergan. In remote sensing. if the light of two or more wavelengths is emitted. the exitance curves of black bodies at various temperatures can be used to model naturally occurring phenomena like solar radiation and terrestrial emmitance. in either case. but. For ETM+ data. color. black body. measurements pertaining to the spectral or geometric characteristics of a sensor or radiation source. the color will be mixed. a number in the range of 0-255 that is related to the amount of planetary radiance striking a sensor's detector. An ideal body which. color composite. Chanduri. reflecting none. That property of an object. transmitters and clock signal typically. The noise associated with periodic signals arising from power supplies. Calibration data are obtained through the use of a fixed energy source such as a calibration lamp. cubic convolution. would be a perfect absorber and a perfect radiator.BCH. C calibration data. brightness value. coherent noise. if blue is assigned to band 1. a true color image will result. the wavelength of the light it emits. If. and red assigned to band 3. White light is a balanced mixture of all the visible spectral colors.

detector sample. Digital information produced by DMA. Data and command transfers from the MOC to the onboard computer. dark shutter image data.e. A NASA requirement to ensure that Landsat 7 data are compatible to those obtained by earlier Landsat satellites. data continuity. data granule. dwell time. The increment of image data stored in the archive. digital terrain elevation data (DTED). The point at which the orbit of an earth satellite intersects the plane of the equator going from north to south. A change in scale from one part of an image to another. or allowed to. The composite circuitry supporting the development of a single output data sample. dynamic range. swath. detector. which provides a uniform matrix of terrain elevation values. Refers to the momentary time interval during which a detector is able to. The process of determining the transfer characteristics (detector mean output as a function of incident exposure) for each detector element.D data capture. distortion. DTED is commonly used to terrain correct Landsat data. or WRS scene. The receipt and storage of return link mission data at the CADU level. data loads. The ratio of the maximum signal to the smallest measurable signal. i.   169  . The image data obtained from ETM+ detectors when the calibration shutter obscures the detectors from incident electromagnetic radiation. sense incoming electromagnetic radiation within its intended instantaneous field of view. an interval. descending node.

emission. ETM+ scene.E EDC. engineering data. ephemeris. distribution and research facility for remotely sensed data and other geographic information. emissivity. May be expressed as total emissivity (for all wavelengths). A set of ETM+ observations that covers 170 km in width by 185 km in length and is centered about a WRS vertex. usually mean sea level. safety. Energy emitted as result of changes in atomic and molecular energy states and propagated through space at the speed of light. Earth Resources Observation System Data Center is a national archive. electromagnetic spectrum. at the constant velocity of light. according to wavelength. (see EROS) electromagnetic radiation. to a point or object on the Earth's surface. Vertical distance from the datum.   170  . The viewing swath is produced by means of an oscillating mirror system that sweeps across track as the sensor field of view moves forward along-track due to satellite motion. The ETM+ is a fixed-position nadir viewing whisk-broom instrument. The system that classifies. Ratio of radiation emitted by a surface to the radiation emitted by a black body at the same temperature under similar conditions. Enhanced Thematic Mapper Plus (ETM+). the process by which a body emits electromagnetic radiation as a consequence of its kinetic temperature only. With respect to electromagnetic radiation. harmonically. environment or status of the platform and instruments. All data available on-board about health. or goniometric emissivity (as a function of angle). production. spectral emissivity (as a function of wavelength). Ephemeris data help to characterize the conditions under which remote sensing data are collected and may be used to correct the sensor data prior to analysis. elevation. A set of data that provides the assigned places of a celestial body (including a manmade satellite) for regular intervals. all energy (from short cosmic to long radio) that moves.

a frame is one Virtual Channel Data Unit with a frame synchronizer pattern (frame marker) attached. focal plane.   171  . geosynchronous. the Earth) with respect to the planes of the geodetic equator and of a reference meridian. Geological Survey. Commonly expressed in terms of latitude and longitude. The solid angle through which an instrument is sensitive to radiation.S. to receive. geodetic accuracy. the plane occupied by the detectors. such as Landsat data. Any coordinate frame whose origin is relative to the Earth's center of mass. The Earth Resources Observation System was established in the early 1970?s under the Department of Interior? U. values and data that have been geometrically corrected using precise ground control point data and elevations models. For Landsat 7. and on which the radiances sensed are incident. focal length. process and distribute data from the United States? Landsat satellite sensors and from airborne mapping cameras. instantaneous field of view. and detector-to-detector sampling delay. geometric correction. Quantities. The transformation of image data. and resolution. A measure of how closely a point on the Earth can be located relative to its true absolute location. to match spatial relationships as they are on the Earth. Includes correction for band-to-band offsets. line length. which define the position of a point on the spheroid of reference (for example. a distinction is made between data that have been geometrically corrected using systematic. Earth rotation. In a camera. For ETM+ data. G geocentric. F field-of-view. This is the same as a Channel Access Data Unit (CADU). See effective resolution element. An Earth satellite orbit in which the satellite remains in a fixed position over a geographic location on Earth. geodetic coordinates. or predicted. frame.EROS. In a sensor. the distance measured along the optical axis from the optical center of the lens to the plane at which the image of a very distant object is brought into focus.

The attribute of a color that differentiates it from gray of the same brilliance and that allows it to be classed as blue. which are acquired from field checks. The vertical projection of the actual flight path of a plane or space vehicle onto the surface of the Earth. or status of the platform and instruments. safety. Any one of a group of operations. image restoration. image smoothing. These operations include contrast enhancement. electro-optical. spatial filtering. A process by which a degraded image is restored to its original condition. or intermediate shades of these colors. optical-mechanical.   172  .Global Position System (GPS). ground control point (GCP). red. The recorded representation of an object produced by optical. environment. Ground truth is used as the basis for making decisions on training areas and evaluating classification results. The registration between images taken at different times. A constellation of satellites that can be used to determine accurately the orbit data of satellites. H housekeeping data. high-resolution remote sensing data. ground track. image-to-image registration. It is the term generally used when the electromagnetic radiation emitted or reflected from a scene is not directly recorded on photographic film. Data. edge enhancement. which improves the interpretability of an image or the detectability of targets or categories in the image. ground truth. or other sources of known data. A geographic feature of known location that is recognizable on images and can be used to determine geometric correction functions for those images. All data available onboard about health. Image restoration is possible only to the extent that the degradation transform is mathematically invertible. hue. and image sharpening. image enhancement. or electronic means. I image. green.

interval.5 to 36 gigahertz. the infrared wavelengths are often subdivided into near infrared (0. Any Landsat ground station not belonging to the United States. Jitter effects arising from the oscillating mirrors and other movable parts aboard the Landsat spacecraft are often a cause of certain anomalies in the image data received and must be compensated for by the ground processing software. The measure. international ground station (IGS). J jitter. Hadamard. are examples of frequently used image transformation procedures. For remote sensing.7 .15 µm). The solid angle through which a detector is sensitive to radiation. of radiant flux incident on a surface. In a scanning system this refers to the solid angle subtended by the detector when the scanning motion is stopped.7 . irradiance. K K-band. Fourier. A radio frequency band extending from approximately 12.image transformation. A function or operator. Depending on the transform chosen. IFOV also refers to the ground area covered by this solid angle. Small rapid variations in a variable (such as a waveform) due to deliberate or accidental electrical or mechanical disturbances or to changes in the supply of voltages.   173  .3-3.1. Far infrared is sometimes referred to as thermal or emissive infrared. middle infrared (1. the input and output images may appear entirely different and have different interpretations. and Karhunen-Love transforms as well as various spatial filters. which takes an image as input and produces an image as its output. in the characteristics of components. in units of power.100 µm wavelength region of the electromagnetic spectrum. and far infrared (7. Pertaining to energy in the 0. and may be from 1 to 90 full scenes in length. instantaneous field of view (IFOV).0 .3 µm). infrared. Is a scheduled ETM+ image period along a WRS path. Instantaneous field of view is commonly expressed in milliradians.0 µm).

An array of values from which functions corresponding to a given argument can be obtained. sun angles.g. In the spatial domain. and other quality measurements. which is the most fundamental element of the data stream structure. minor frame. which can be used in the convolution or multiplication with a NxN image to accentuate certain features or properties of an image. level 0. transmitted.kernel. metadata. Consists of the spacecraft and the ETM+ payload. L L-band. map projection. For ETM+ major frames are partitioned into minor frames. a major frame period is one complete scan of the ETM+ scan mirror (either direction). imagery. long term acquisition plan. Space vehicle or instrument data at full space-time resolution with space-to-ground communication artifacts removed. Light that has traveled through a medium without being absorbed or scattered. lookup table. light.   174  . A radio frequency band extending from approximately 1. which includes not only the period during a scan but also the turnaround interval when the scan mirror changes direction for the next scan. date of acquisition. a kernel is a MxM operator. PCD.0 gigahertz. in which specific data measurands (e. gain states.0 to 2. M major frame. could cover. Landsat 7. An archived set of descriptive information about a scene and the parent subinterval that provides a user with geographic coverage. A kernel can also be represented in the frequency domain as a Fourier transform. The tasking of the sensor using cloud predictions to optimize the acquisition of cloud free scenes. For ETM+. time codes) are extracted. Any systematic arrangement of meridians and parallels portraying the curved surface of a sphere or spheroid upon a plane.

That point on the celestial sphere vertically below the observer. or control. or amplitude of an electromagnetic wave or other variable. and any other data which is required to perform mirror scan correction. modulation transfer function (MTF). The preferred term for the shorter wavelengths in the infrared region extending from about 0.mirror scan correction data. To vary. multiplexer. narrowband data. As an imaging system processes or records an image. the contrast modulation of the processed or recorded image is different from the input image. Generally denotes remote sensing in two or more spectral bands. multispectral. An image made by piecing together individual images covering adjacent areas. The data includes the command or forward ranging in the narrowband forward link. phase. near infrared. which permits the transmission of multiple messages simultaneously on one communication channel. The MTF can be thought of as a curve. An electronic device. scan direction. This data includes scan start time. Sometimes called solar infrared. The modulation transfer function of an imaging system measures the spatial frequency modulation response of the system. as it is only available for use during the daylight hours. The longer wavelength end grades into the middle infrared. N nadir. second half scan time error. indicating for each spatial frequency the ratio of the contrast modulation of the output image to the contrast modulation of the input image. first half scan time error. such as visible and infrared. the frequency. Also known as the shortwave infrared (SWIR). It is formally defined as the magnitude of the Fourier transform of the line spread function of the imaging system. mosaic. modulate.7 µm (visible red) to about 3 µm.   175  . and the telemetry or return ranging in the narrowband return link. or 180° from the zenith.

orthorectified. track. noise. parallax. Nyquist's theorem: A theorem. Nyquist. developed by H. Nyquist interval. the output is obtained. The apparent change in the position of one object.   176  . The Nyquist interval is equal to the reciprocal of twice the highest frequency component of the sampled signal. Given any two of these three entities. The adding to or taking away of orbital velocity. Any unwanted disturbance affecting a measurement (as of a frequency band). A single band covering a broad range of wavelengths. Describing an image in which terrain relief distortions have been removed. especially that which degrades the information-bearing quality of the data of interest. The maximum time interval between equally spaced samples of a signal that will enable the signal waveform to be completely determined.node. orbital period. twice the highest frequency component in the analog signal. the orbital nodes occur at the equator. one on the descending. Either of the two points at which the orbit of a heavenly body intersects a given plane. the third can be obtained. or daylight. which states than an analog signal waveform may be uniquely reconstructed. or point. When the transfer function operates on the input. orbit adjust. The interval in time between successive passages (orbits) of a satellite through a reference plane. The sampling rate must be equal to. With respect to Landsat. A mathematical statement that describes the relationship between the input and the output of an imaging system. O optical transfer function (OTF). P panchromatic. with respect to another. from samples taken at equal time intervals. without error. or nighttime. This is normally done to maintain altitude or orbit phasing relationships. especially the plane of ecliptic. or greater than. track of the orbit and the other on the ascending. usually used in context of collecting information from the whole visible spectrum. when viewed from different angles.

path. The longitudinal center line of a Landsat scene of a Landsat scene, corresponding to the center of an orbital track. Sequential numbers from east to west are assigned to 233 nominal satellite tracks for Landsat 7. Path numbers are used with row numbers to designate nominal scene center points. payload. That part of a spacecraft (e.g. ETM+) that is separate from the equipment or operations necessary to maintain the spacecraft in orbit. payload correction data. Image support data imbedded in the wideband data stream. Includes satellite attitude, ephemeris, time, angular displacement sensor (ADS) data and payload state. perigee. The point in the orbit of a heavenly body (e.g. satellite) at which it is nearest the Earth. pixel. Picture element provided by a single detector scene sample output. pitch. The rotation of a spacecraft about the horizontal axis normal to its longitudinal axis (in the along-track direction) so as to cause a nose-up or nose-down attitude. polar stereographic. An azimuthal stereographic projection commonly used with Landsat data acquired about 65° latitude. In this projection, the meridians are straight lines converging at the pole (central point), and lines of latitude are concentric circles about this point. Like the UTM projection, the polar stereographic is a conformal projection, meaning that angular relationships are preserved. pole wander. The apparent motion in the poles of the Earth relative to inertial coordinate system. Changes in moments of inertia are due to changes in moments of density due primarily to tides and liquid mass. The National Imager and Mapping Agency (NIMA) generates pole wander data which are used by the Landsat 7 system in the conversion of downlinked ephemeris from inertial to fixed reference, during Level 0R processing. precision correction. Post-processed geometric correction of satellite data using ground control points to correlate the spacecraft's predicted position with its actual geodetic position. prime meridian. Meridian of longitude 0 degrees, used as the origin for measurements of longitude. The meridian of Greenwich, England, is the internationally accepted prime meridian on most charts.   177 

Q quantization level. The number of numerical values used to represent a continuous quantity. quaternion. A vector of four components; the position is contained in the first three components and an associated scalar rater is located in the last component of this four element vector. R radian. The angle subtended by an arc of a circle equal in length to the radius of the circle: 57.3° radiance. Measure of the energy radiated by an object. In general, radiance is a function of viewing angle and spectral wavelength and is expressed as energy per solid angle. Rayleigh scattering. Selective scattering of light in the atmosphere by particles that are small compared with the wavelength of light. reflectance. The ratio of the radiant energy reflected by a body to that incident upon it. In general, reflectance is a function of the incident angle of the energy, viewing angle of the sensor, spectral wavelength and bandwidth, and the nature of the object. registration. The process of geometrically aligning two or more sets of image data such that resolution cells for a common ground area can be digitally or visually superimposed. roll. The rotation of a spacecraft about its longitudinal axis (in the along-track direction) so as to cause a side-up or side-down attitude. The roll axis is referred to as the y axis. row. The latitudinal (nominal) center line of a Landsat scene. Row 1 is at latitude 80° 47'N, row 60 is at the equator, and row 122 is at latitude 81° 51'S. In total there are 248 rows. S sampling rate. The number of samples taken per unit time, i.e., the rate at which signals are sampled for subsequent use, such as for modulation, coding, and quantization. saturation. The condition where energy flux exceeds the sensitivity range of a detector. S band. A radio frequency band extending from approximately 2.0 to 4.0 gigahertz.   178 

sidelap. The extent of lateral overlap between images acquired over adjacent ground tracks. signal-to-noise ratio. The ratio of the level of the information-bearing signal power to the level of the noise power. More precisely, the signal-to-noise ratio of the mean DN to the standard deviation in DN. This is a temporal noise definition in that the mean DN is the time averaged value and the standard deviation in DN is the standard deviation in the time series. space oblique mercator. A variation on the basic mercator map projection based on the dynamics of satellite motion. The movements of the satellite, sensor, and the Earth, expressed as functions of time, are used to calculate which latitudes and longitudes on the Earth correspond to locations in the projection plane. spectral band. An interval in the electromagnetic spectrum defined by two wavelengths, frequencies, or wave numbers. spectral response. The response of a material as a function of wavelength to incident electromagnetic energy, particularly in terms of the measurable energy reflected from and emitted by the material. spectral signature. The quantitative measurement of the properties of an object at one or several wavelength intervals. Spectral signature analysis techniques use the variation in the spectral reflectance or emittance of objects as a method of identifying the objects. steradian. A unit of measure of solid angles. Formally, it is the angle subtended at the center of the sphere by a portion of the surface whose area is equal to the square of the radius of the sphere. There are 4 pi steradians in a sphere. subinterval. Is a contiguous segment of raw wideband data received during a Landsat 7 contact period. Subintervals are caused by breaks in the wideband datastream due to communication dropouts and/or the inability of the spacecraft to transmit a complete observation (interval) within a single Landsat 7 contact period. The largest possible subinterval is 35 full scenes long with a partial scene preamble and postamble. The smallest possible subinterval is a single ETM+ scene. sun elevation angle. The angle of the Sun above the horizon. solar zenith angle. Reciprocal of the sun elevation angle.

 

179 

This will be mean Sun time relative to Greenwich. Abbreviated GMT. In practice the limits represent the envelope of energy emitted by the Earth behaving as a graybody with a surface temperature around 290 K. Mean solar time of the meridian of Greenwich. The thermal band most used in remote sensing extends from 8 to 15 µm. Refers to the 185 kilometer wide ETM+ imaging ground track. concerned with. moving east) by 15 and adding the result to the current GMT. thermal band. mean Sun. used by most navigators and adopted as the prime basis of standard time throughout the world. Pertaining to. A general term for intermediate and long wavelength infrared-emitted radiation. precessing through 360° during the period of a year. The mean Sun time at any location is determined by dividing the difference in longitude from Greenwich (in degrees. the radiance envelope has several brighter bands corresponding to windows in the atmospheric absorption bands. The distance in longitude from the Greenwich meridian determines the mean Sun time at a given location on the Earth. or limited by time. and there. to about 15 or 20 m where the far infrared commences. thermal infrared. In practice. expressed in hours. temporal. interpreting or recording the quantity measured. time.   180  .sun synchronous. swath. An Earth satellite orbit in which the orbital plane remains at a fixed angle with respect to the Sun. time. Greenwich mean. generally refers to infrared radiation emitted in the 3-5 µm and 9-14 µm atmospheric windows. as contrasted to short wavelength reflected infrared radiation. England (longitude 0). T telemetry. The science of measuring a quantity. Seen from space. The expected repeat time between measurements over the same location. The preferred term for the middle wavelength ranges of the infrared region extending roughly from 3 µm at the end of the near infrared. transmitting the measured value to a distant station. temporal resolution.

each covering 6 degrees of longitude and oriented to a meridian. In general. A widely used map projection employing a series of identical projections around the world in the intermediate latitudes. and by the ease with which it allows a useful rectangular grid to be superimposed on it. spectral wavelength and bandwidth. UT1-UTC time correction data. Universal Time (UT) 1 is determined from observations of stellar transits to determine local mean sidereal time corrected to remove the effects of polar motion. umbra.4 to 0. V virtual channel data unit (VCDU). meaning that it preserves scale and angular relationships well. Electromagnetic radiation of shorter wavelength than visible radiation but longer than X-rays.transmittance.9 seconds of UT1 by periodic leap-second adjustments.000 angstroms. the spectral interval from approximately 0. roughly. and the nature of the object. transmittance is a function of the incident angle of the energy. but it is kept with . The ratio of the energy per unit time per unit area (radiant power density) transmitted through an object to the energy per unit time per unit area incident on the object.7 µm. universal transverse mercator. where the light from the source of illumination is completely cut off. as a planet. radiation in the wavelength interval between 10 and 4. Electromagnetic radiation of the wavelength interval to which the human eye is sensitive. U ultraviolet radiation. Universal Time Coordinated (UTC) is defined to be equal to that of the International System used for atomic time. The UTM projection is characterized by its property of conformality. The complete or perfect shadow of an opaque body. The UTM projection is most commonly used with landsat data.   181  . visible radiation. The CCSDS protocol data unit consisting of a fixed length data structure. viewing angle of the sensor. It is used for bidirectionally space/ground communications on a CCSDS virtual channel.

A global indexing system for Landsat data.0 to 12.5 gigahertz.                                       182  . In general. world geodetic system (WGS). The point in the celestial sphere that is exactly overhead. Wavelength = 1/frequency. Specifically. The rotation of a spacecraft about its vertical axis so as to cause the spacecraft's longitudinal axis to deviate left or right from the direction of flight. worldwide reference system. X X-band.W wavelength. defined by path and row coordinates. Z zenith. The yaw axis is referred to as the z axis. Y yaw. which is based on nominal scene centers. A radio frequency band extending from approximately 8. the shortest distance between particles moving in the same phase of oscillation in a wave disturbance. the mean distance between maximums (or minimums) of roughly periodic pattern. The reference Earth model used by the Landsat 7 system.

Distributed Active Archive Center Earth Observing System Earth Observation Satellite Company EOS Data and Information System EOS Polar Ground Sites Earth Resources Observation Systems Enhanced Thematic Mapper Enhanced Thematic Mapper Plus Full Aperture Calibrator Full Aperture Solar Calibrator Flight Dynamics Facility Flight Operations Manager Flight Operations Team File Transfer Protocol Geographic Tagged Image File Format Ground Sample Distance Hierarchical Data Format Ion Assisted Deposition Image Assessment System Internal Calibrator Instantaneous Field of View International Ground Stations Initial On-orbit Checkout Infrared Jet Propulsion Laboratory Landsat Ground Network Landsat Ground Station 183  .Acronym Expansion ACCA AGS AKM AU BRDF CADU CCSDS CN COFUR CPF CRAM DFCB DN DOQ ECS EDC EDC-DAAC EOS EOSAT EOSDIS EPGS EROS ETM ETM+ FAC FASC FDF FOM FOT FTP GeoTIFF GSD HDF IAD IAS IC IFOV IGS IOC IR JPL LGN LGS   Automatic Cloud Cover Assessment Alaska Ground Station Apogee Kick Motor Astronomical Unit Bidirectional Reflection Distribution Function Channel Access Data Units Consultative Committee for Space Data Systems Coherent Noise Cost Of Fulfilling User requests Calibration Parameter File Combined Radiometric Correction Model Data Format Control Book Background Noise Digital Orthophoto Quadrangle EOSDIS Core System EROS Data Center EROS Data Center .

                    LP-DAAC LPGS LPS LTAP ME MOC MMS MSCD MSS MTF NASCOM NCEP NISN NLAPS NOAA PAC PCD SBRS SCS SGS SLC SMA SME SNR SSR SWIR SZA TDRS TM TOA USAF USGS VNIR WFF WGS WRS Land Processes Distributed Active Archive Center Level-1 Product Generation System Landsat Processing System Long Term Acquisition Plan Memory Effect Mission Operations Center Multimission Modular Spacecraft Mirror Scan Correction Data MultiSpectral Scanner Modulation Transfer Function NASA Communications Division National Centers for Environmental Prediction NASA Integrated Services Network National Landsat Archive Production System National Oceanic and Atmospheric Administration Partial Solar Calibrator Payload Correction Data Santa Barbara Remote Sensing Scan Correlated Shift Svalbard Ground Station Scan Line Corrector Scan Mirror Assembly Scan Mirror Electronics Signal to Noise Ratio Solid State Recorder Short Wave Infrared Solar Zenith Angle Tracking Data and Relay Satellites Thematic Mapper Top-of-Atmosphere United State Air Force United States Geological Survey Visible & Near Infrared Wallops Flight Facility World Geodetic System Worldwide Reference System 184  .

N.. and Bindschadler.” IEEE Transactions on Geoscience and Remote Sensing. K.. S.  Ellison. 42. and Assimilation of Clouds. vol. Haskett.. European Centre for Medium-Range Weather Forecasts.. Validation. Sylvia S... Meyer..  Goward. Shen. E. (1999). Dec. Use of cloud analyses to validate and improve model diagnostic clouds at NMC. and Goward.  Chander. Turner.. Reeley.. 31Nov. 2821-2831. Irons. Milstein. Ruggles. 1994.  Campana.. “Landsat-5 TM Reflective-Band Absolute Radiometric Calibration”. Markham. England. (2000). J. no. ETM+. J. R. “Summary of current radiometric calibration coefficients for Landsat MSS. pp. Thome. vol. B. Hyperspectral. (1995).. 4049. Oct. A. “Revised Landsat 5 Thematic Mapper Radiometric Calibration. Available in PDF form. in press Remote Sensing of the Environment. S. 2747-2760. Enhanced Landsat capturing all the Earth's land   185  . J. G. “Cross-Calibration of the Landsat-7 ETM+ and EO-1 ALI sensors.L. D. 4. Williams.. pp. T. Dubayah.. Descour.L. Jul.  Chander. pp. 490–494.. Cloud cover avoidance in space-based remote sensing acquisition.. D.. No. A. Available in PDF form. Campana.. R. (2001). Editors. 42. (Invited paper) Available in PDF form. (1994). K. J.L.. Michael R. 4.. B. Lonigro..An innovative approach to building a global imagery archive. ” IEEE Transactions on Geoscience and Remote Sensing. Micijevic.  Chander. vol. Kaita. J. Gasch. D. Helder.. Dewald.  Gasch. and EO-1 ALI sensors”.. 2007. Available in PDF form. N. 3.. 336-347. J. 12. Landsat 7's Long Term Acquisition Plan . (Invited paper) Available in PDF form. J. Dec.L. K. Remote Sensing of Environment 113 (2009) 893 – 903. and Campana... T. Gasch. G. R. E. M. L. Helder. B.References  Arvidson.” IEEE Transactions on Geoscience and Remote Sensing. G. and Ultraspectral Imagery IV.. Markham. A. The Aerospace Corporation.J..J. 12. G. pp. Markham. D. In Algorithms for Multispectral. 2004. K.. T. No. Paper presented at the Joint ECMWF/GEWEX Workshop on Modelling. Improved Reduced-Resolution Satellite Imagery.  Chander.. Reading. Helder. J. Proceedings of SPIE Vol. S. J. TM. Arvidson..L. Barsi.. 2004. D.

(1996). (2000)..A. S. B. B.  Irish. EOS. December 2001. Barker. J. 348-355.. and Next-Generation Satellites V. J. J. 72. and Ultraspectral Imagery. Early Ground-Reference Calibration Results for Landsat-7 ETM+ Using Small Test Sites. Available in PDF form. Walker. Geneva. L.. 115 pp. Available in PDF form. Raqueno. pp.J. and Nandy. (2000)..    186  . pp. Remote Sensing of Environment. W. K. Switzerland... Proceedings of SPIE -.. Photogrammetric Engineering & Remote Sensing. Joan B. Chander... "Landsat 7 on-orbit modulation transfer function estimation". J. Landsat 7 Automatic Cloud Cover Assessment. Nordgren. SPIE Conference. Barker. no. 10. E. Barsi. Proc. B. Sensors.. pp.  Irish. and Morfitt. and de Alwis. Beuschel.R. August 1986. Arvidson (2006).. Seiferth. James. Calibration of Landsat Thermal Data and Application to Water Resource Studies.G.. L. No. 1.  Rossow.  Markham. J.  Thome. Available in PDF form.  Markham. B. World Meteorol. SPIE Vol. Available in PDF form. Available in PDF form. M. and T.L. C. P. Vol. in press. R. A. L. D. Hyperspectral. Characterization of the Landsat-7 ETM+ automated cloud-cover assessment (ACCA) algorithm..areas. WMO/TD-No. 134-142... Konradin Weber. Anderson.J. On-orbit Performance of the Landsat-7 ETM+ Radiometric Calibrators. R. R. Exoatmospheric Reflectances and At-Satellite Temperatures. W. vol.L. Goward.R. LaMarr. White Paper. D. B. 737. Landsat MSS and TM Post-Calibration Dynamic Ranges. 50-61. 80 (26): 289.L.Volume 4540. Lurie. J.. 4049. Whittington.  Markham. N. and Roiter. Organ. pp: 1179–1188. 2003. N.L. in press. Algorithms for Multispectral. 293. International Journal of Remote Sensing. 1996  Schott. Systems. E.N. International Satellite Cloud Climatology Project (ISCCP) Documentation of New Cloud Datasets... EOSAT Landsat Technical Notes. D. Editors. G.  Storey. 4049. Barker. Revised Landsat 5 TM Radiometric Calibration Procedures and Post-Calibration Dynamic Ranges.. Hiroyuki Fujisada. J. May..

Sign up to vote on this title
UsefulNot useful

Master Your Semester with Scribd & The New York Times

Special offer for students: Only $4.99/month.

Master Your Semester with a Special Offer from Scribd & The New York Times

Cancel anytime.